Laboratory techniques in biochemistry and molecular biology
6
LABORATORY TECHNIQUES IN BIOCHEMISTRY AND MOLECULAR BI...
43 downloads
1103 Views
21MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Laboratory techniques in biochemistry and molecular biology
6
LABORATORY TECHNIQUES IN BIOCHEMISTRY AND MOLECULAR BIOLOGY Volume 6 Edited by
T.S. WORK - N.I.M.R.,Mill Hill, London E. WORK - ‘East Lepe’, 60 Solent View Road, Cowes, Isle of Wight
Advisory board G. POPJAK - U.C.L.A. S . BERGSTROM - Stockholm K. BLOCH - Harvard University P . SIEKEVITZ - Rockefeller University E. SMITH - U.C.L.A. E.C. SLATER - Amsterdam
NORTH-HOLLAND PUBLISHING COMPANY AMSTERDAM . NEW YORK . OXFORD
Part I
Part I1
R. Hinton and M . Dobrota DENSITY GRADIENT CENTRIFUGATION T. Chard A N INTRODUCTION TO RADIOIMMUNOASSAY A N D RELATED TECHNIQUES
1978 NORTH-HOLLAND PUBLISHING COMPANY AMSTERDAM . N E W YORK.OXFORD
0 ElsevierfNorth-Holland Biomedical Press, 1978 AN rights reserved. No parts of this publication may be reproduced, stored in a retrieval system. or transmitted, in any form or by any means, electronic. mechanical, photocopying. recording or otherwise, without the prior permission of the copyright owner.
ISBN - series: 072044200 I - volume 6 : 072044221 4
Published by: ELSEVIERINORTH-HOLLAND BIOMEDICAL PRESS 335 JAN VAN CALENSTRAAT, P.O.BOX 21 1 AMSTERDAM, THE NETHERLANDS
Sole disrributorsjor the U.S.A. and Canada: ELSEVIER/NORTH-HOLLAND INC 52 VANDERBILT AVENUE NEW YORK. N.Y. 10017
Printed in The Netherlands
Editors’ preface
Progress in research depends upon development of technique. No matter how important the cerebral element may be in the planning of experiments, a tentative hypothesis cannot be converted into an accepted fact unless there is adequate consciousness of the scope and limitation of existing techniques; moreover, the results may be meaningless or even positively misleading if the technical ‘know how’ is inadequate. During the past ten or fifteen years, biochemical methods have become specialized and sophisticated to such a degree that it is now difficult for the beginner, whether undergraduate, graduate or specialist in another field, to grasp all the minor but important details which divide the successful from the unsuccessful experiment. In order to cope with this problem, we have initiated a new series of Laboratory Manuals on technique. Each manual is written by an expert and is designed as a laboratory handbook to be used at the bench. It is hoped that use of these manuals will substantially reduce or perhaps even remove that period of frustration which so often precedes the successful transplant of a specialized technique into a new environment. In furtherance of this aim, we have asked authors to place special emphasis on application rather than on theory; nevertheless, each manual carries sufficient history and theory to give perspective. The publication of library volumes followed by pocket paperbacks is an innovation in scientific publishing which should assist in bringing these manuals into the laboratory as well as into the library. In undertaking the editing of such a diverse series, we have become painfully conscious of our own ignorance but have been enV
VI
EDITORS PREFACE
couraged by our board of advisers to whom we owe many valuable suggestions and, of course, by our authors who have co-operated so willingly and have so patiently tolerated our editorial intervention.
T. S. & E. Work Editors
Contents of parts I and I1
PART I DENSITY GRADIENT CENTRIFUGATION. Richard Hinton and Miloslav Dobrota Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter I . Introduction to zonal centrifugation . . . . . . . . . . . . . Chapter 2 . Theoretical aspects of centrifugal separations . . . . . . . . . . Chapter 3. Conditions for a centrifugal separation . . . . . . . . . . . . . Chapter 4 . Centrifugation in conventional rotors . . . . . . . . . . . . . . Chapier 5 . Centrifugation in zonal rotors . . . . . . . . . . . . . . . . . Chapter 6. Assay of fractions separated by density gradient centrifugation . . . Chapter 7. Applications of density gradient centrifugation . . . . . . . . . Chapter 8. Artifacts arising during centrifugal separations . . . . . . . . . . Chapter 9. Future prospects for density gradient centrifugation . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix I1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 111 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
3 8 46 70 97 120 197 205 243 254 262 263 265 269 271 274 287
PART I1 AN INTRODUCTION TO RADIOIMMUNOASSAY AND RELATED TECHNIQUFS. T. Chard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 299 List of abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . Chapter I . The background to radioimmunoassay . . . . . . . . . . . . . 301 Chaprer 2. Requirements for a binding assay - purified ligand . . . . . . . . 329 Chapier 3. Requirements for binding assays - tracer ligand . . . . . . . . . 343 VII
Vlll
CONTENTS OF PARTS I AND 11
Chapter 4. Requirements for a binding assay .the binder . . . . . . . . . . Chapter 5. Requirements for a binding assay .separation of bound and free ligand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter 6. Requirements for a binding assay extraction of ligand from biological fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter 7. Requirements for binding assays - calculation of results . . . . . . Chapter 8. Characteristics of binding assays - sensitivity . . . . . . . . . . Chapter 9. Characteristics of binding assays - specificity . . . . . . . . . . Chapter 10. Characteristics of binding assays - precision . . . . . . . . . . Chaprer 11. Characteristics of binding assays - relation to other types of assay . Chapter 12.Automation of binding assays . . . . . . . . . . . . . . . . Chapter 13. Organisation of assay services . . . . . . . . . . . . . . . . Appendix I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix I1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix I11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
311 401
~
427 440 446 463 419 49 5 504 510 518 520 521 522 523 521
531
DENSITY GRADIENT CENTRIFUGATION Richard Hinton and Miloslav Dobrota Wol/son Bioanalytical Centre, University o f Surrey,
Guildford, Surrey GU2 S X H , U.K.
This Page Intentionally Left Blank
Contents
Chapter 1 . Introduction to zonal centrifugation . . . . . . . .
8
1.1. The first applications of centrifugation in biology . . . . . . . . . . . 1.2. Centrifugal techniques . . . . . . . . . . . . . . . . . . . . . . 1.2.1. Analytical ultracentrifugation . . . . . . . . . . . . . . . . . . 1.2.2. Differential pelleting . . . . . . . . . . . . . . . . . . . . . 1.2.3. Rate-zonal centrifugation . . . . . . . . . . . . . . . . . . . 1.2.4. lsopycnic zonal centrifugation . . . . . . . . . . . . . . . . . 1.3. The development of centrifuges and rotors . . . . . . . . . . . . . . 1.3.1. The centrifuge . . . . . . . . . . . . . . . . . . . . . . . . 1.3.2. Rotor materials . . . . . . . . . . . . . . . . . . . . . . . . 1.3.3. Rotorshape . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.4. Zonal rotors . . . . . . . . . . . . . . . . . . . . . . . . . 1.4. Uses and limitations of centrifugal techniques . . . . . . . . . . . . . 1.5. Design of a centrifuge laboratory . . . . . . . . . . . . . . . . . . 1.6. Safety ofcentrifuges . . . . . . . . . . . . . . . . . . . . . . . . 1.6.1. Causes and prevention of rotor failure . . . . . . . . . . . . . . 1.6.I .1 . Overspeeding . . . . . . . . . . . . . . . . . . . . . . . 1.6.1.2. Corrosion . . . . . . . . . . . . . . . . . . . . . . . . 1.6.1.3. Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . 1.6.1.4. High gradient densities . . . . . . . . . . . . . . . . . . . 1.6.1.5. Vacuum failure . . . . . . . . . . . . . . . . . . . . . . 1.6.1.6. Freezing of rotor contents . . . . . . . . . . . . . . . . . . 1.6.2. Other precautions in operation of centrifuges . . . . . . . . . . . 1.6.2.1. Access to spinning rotor . . . . . . . . . . . . . . . . . . 1.6.2.2. Aerosol formation . . . . . . . . . . . . . . . . . . . . . 1.6.2.3. Electrical safety . . . . . . . . . . . . . . . . . . . . . . 1.6.2.4. Balancing and assembly of rotors . . . . . . . . . . . . . . 1.6.2.5. Compatability of centrifuge and rotor . . . . . . . . . . . . . 1.6.2.6. Water leaks . . . . . . . . . . . . . . . . . . . . . . . 1.6.2.7. Flammable solvents . . . . . . . . . . . . . . . . . . . .
8 9 9 11 13 15 16 16
3
18
19 20 24 27 29 29 30 32 32 33 33 33 34 34 34 35 35 35 35 36
4
DENSITY GRADIENT CENTRIFUGATION
1.7. Care of rotors . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7.1. Choice of rotor . . . . . . . . . . . . . . . . . . . . . . . . . 1.7.2. Properties and care of materials used in rotor construction . . . . . . 1.7.3. Washing and drying rotors . . . . . . . . . . . . . . . . . . . 1.7.4. Assembly, disassembly and storage . . . . . . . . . . . . . . . 1.7.5. Care of individual rotors . . . . . . . . . . . . . . . . . . . . 1.8. Guarantees on rotors . . . . . . . . . . . . . . . . . . . . . . .
36 36 36 40 41 41 43
Chapter 2 . Theoretical aspects of centrifugal separations
46
. . . .
2.1. Theory of rate-zonal separations . . . . . . . . . . . . . . . . . 2.1.1. Theory of sedimenting particles . . . . . . . . . . . . . . . . 2.1.2. Stability of the sample zone . . . . . . . . . . . . . . . . . . 2.1.3. Stability of a sedimenting zone . . . . . . . . . . . . . . . . 2.2. Theory of isopycnic banding . . . . . . . . . . . . . . . . . . . . 2.2.1. Shape of zones separated by isopycnic banding . . . . . . . . . 2.2.2. Redistribution of density gradient solutes in a centrifugal field . . . 2.2.3. Influence of the rotor on gradient shape . . . . . . . . . . . . 2.3. Effects of density gradient solutes on subcellular structures . . . . . .
Chapter 3 . Conditions for a centrifugal separation
. . .
. . .
46 47 51 57 59 59 60 64 64
. . . . . .
70
. .
3.1. Choice of approach . . . . . . . . . . . . . . . . . . . . . . . . 70 3.2. Choiceofrotor . . . . . . . . . . . . . . . . . . . . . . . . . . 75 3.3. Density gradient solutes . . . . . . . . . . . . . . . . . . . . . . 79 3.3.1. Salts of alkali metals . . . . . . . . . . . . . . . . . . . . . . 80 3.3.2. Small hydrophilic organic molecules . . . . . . . . . . . . . . . 81 3.3.3. High molecular weight organic compounds . . . . . . . . . . . . 86 3.3.4. Other types of density gradient solutes . . . . . . . . . . . . . . 87 3.3.5. Choice of gradient material . . . . . . . . . . . . . . . . . . . 89 3.4. Choice of gradient . . . . . . . . . . . . . . . . . . . . . . . . 90 3.4.1. Gradient for rate-zonal separations . . . . . . . . . . . . . . . 90 3.4.2. Gradient for isopycnic separations . . . . . . . . . . . . . . . 91 3.4.3. Design of complex gradients . . . . . . . . . . . . . . . . . . 93
Chapter 4 . Centrifugation in conventional rotors 4.1. Rate-zonal centrifugation . . . . . . . . . . 4.1 .1. Preparation of the density gradient . . .
. . . . . . .
97
........... . . . . . . . . . . . .
97 97
CONTENTS
4.1.2. Layering of sample on to the gradient . . . . . . . . . . . . . . 4.1.3. Centrifugation . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.4. Recovery from the gradient . . . . . . . . . . . . . . . . . . 4.1.5. Monitoring the displaced gradient . . . . . . . . . . . . . . . . 4.2. Isopycnic zonal centrifugation . . . . . . . . . . . . . . . . . . . 4.2.1. Preparation of the density gradient . . . . . . . . . . . . . . . 4.2.2. Layering of sample on to the gradient . . . . . . . . . . . . . . 4.2.3. Centrifugation . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4. Displacement and monitoring of the gradient . . . . . . . . . . .
5 103 105 108 112 115 115 117 118 118
Chapter 5 . Centrifugation in zonal rotors . . . . . . . . . . . . 120 5 . I . Conventional. non-reorienting zonal rotors . . . . . . . . . . . . . . 5.1.1. Construction of rotors . . . . . . . . . . . . . . . . . . . . . 5.1.2. Types of zonal rotors . . . . . . . . . . . . . . . . . . . . . 5.1.2.1. A-XI1 rotor . . . . . . . . . . . . . . . . . . . . . . . 5.1.2.2. Z-15rotor . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2.3. HS rotor . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2.4. B-IVrotor . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2.5. B-XIV and B-XV rotors . . . . . . . . . . . . . . . . . . 5.1.2.6. B-XXIX and B-XXX rotors . . . . . . . . . . . . . . . . 5.1.2.7. Interchangeable batch and continuous-flow rotors . . . . . . . 5.1.2.8. Continuous-flow rotors . . . . . . . . . . . . . . . . . . . 5.1.2.9. Other rotors . . . . . . . . . . . . . . . . . . . . . . . 5.1.3. Operation of zonal rotors . . . . . . . . . . . . . . . . . . . 5.1.3.1. Preparations for a run . . . . . . . . . . . . . . . . . . . 5.1.3.2. Gradient loading . . . . . . . . . . . . . . . . . . . . . . 5.1.3.3. Sample loading . . . . . . . . . . . . . . . . . . . . . . 5.1.3.4. Acceleration spin and deceleration . . . . . . . . . . . . . . 5.1.3.5. Unloading . . . . . . . . . . . . . . . . . . . . . . . . 5.1.3.6. Monitoring and fraction collecting . . . . . . . . . . . . . . 5.1.4. Ancillary equipment . . . . . . . . . . . . . . . . . . . . . . 5.1.4.1. Gradient makers . . . . . . . . . . . . . . . . . . . . . 5.1.4.2. Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4.3. Refractometers and density meters . . . . . . . . . . . . . . 5.1.4.4. Spectrophotometers, colorimeters and UV meters . . . . . . . 5.1.4.5. Flow-through cells . . . . . . . . . . . . . . . . . . . . 5.1.4.6. Thermometers . . . . . . . . . . . . . . . . . . . . . . 5.1.4.7. Perfusor pumps . . . . . . . . . . . . . . . . . . . . . . 5.1.4.8. Stroboscopic lamps . . . . . . . . . . . . . . . . . . . . 5.1.4.9. Recorders . . . . . . . . . . . . . . . . . . . . . . . .
122 122 130 130 134 134 135 136 139 140 141 142 143 144 153 156 163 168 174 180 180
182 183 185 185 186 186 186 186
6
DENSITY GRADIENT CENTRIFLIGATION
5.1.4.10. Cooling equipment . . . . . . . . . . . . . . . . . . . . . 5.1.4.11. Minor accessories . . . . . . . . . . . . . . . . . . . . 5.1.4.12. Fraction collectors . . . . . . . . . . . . . . . . . . . . 5.1.4.13. Integrator . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Reorienting zonal rotors . . . . . . . . . . . . . . . . . . . . . .
187 187 188 189 189
Chapter 6 . Assay of fractions separated by density gradient centrifugation . . . . . . . . . . . . . . . . . . . . . . 197 6.1. 6.2. 6.3. 6.4.
Enzyme and chemical assays on fractions . . . . . . . . . . . . . . Electron microscopic examination of fractions . . . . . . . . . . . . Assessment of results from density gradient separations . . . . . . . . Calculation of sedimentation coefficients . . . . . . . . . . . . . . .
197 198 200 202
Chapter 7. Applications of density gradient centrifugation . . . . 205 7.1. Separation of living cells . . . . . . . . . . . . . . . . . . . . . . 7.2. Separation of cell organelles from mammalian tissues . . . . . . . . . 7.2.1. Subfractionation of a crude nuclear fraction and separation of large sheets of plasma membranes . . . . . . . . . . . . . . . . . . 7.2.1.1. Subfractionation of purified nuclei . . . . . . . . . . . . . . 7.2.2. Subfractionation of the mitochondria1 fraction . . . . . . . . . . 7.2.3. Subfractionation of the lysosomal fraction . . . . . . . . . . . . 7.2.4. Subfractionation of microsomes . . . . . . . . . . . . . . . . . 7.2.5. Fractionation of ribonucleoprotein particles . . . . . . . . . . . . 7.2.6. Fractionation of chromatin . . . . . . . . . . . . . . . . . . . 7.3. Separation of subcellular structures from plant cells . . . . . . . . . . 7.4. Separation of subcellular structures from unicellular organisms . . . . . 7.5. Fractionation of macromolecules . . . . . . . . . . . . . . . . . . 7.5.1. Nucleic acids . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.2. Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6. Other applications of density gradient centrifugation in biochemistry . . . 7.6.1. Fractionation of serum lipoproteins . . . . . . . . . . . . . . . 7.6.2. Separation ofviruses . . . . . . . . . . . . . . . . . . . . . . 7.7. Other applications of density gradient centrifugation . . . . . . . . . .
205 208 210 211 212 214 217 222 225 226 229 232 232 235 237 237 239 242
7
CONTENTS
Chapter8 . Artifacts arising during centrijugal separations . . . 243 . . . . . . . . . . . . . . . . . . 8.1.2. Damage due to high hydrostatic pressure . . . . . . . . 8.1.3. Damage due to high concentration of the gradient solute .
8 . I . Damage to particles during centrifugation 8.1.1, Damage caused by pelleting . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
8.2. Factors affecting the accuracy of assays performed on fractions from density gradients . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.1. Reaction of the gradient solute with reagents used in the assay of separated constituents . . . . . . . . . . . . . . . . . . . . . 8.2.2. Interference with the performance of analytical equipment . . . . . 8.2.3. Interference in the sensitivity of assays . . . . . . . . . . . . . . 8.3. Uncertainties in estimates of particle density and sedimentation coefficient 8.3.1. Particle density . . . . . . . . . . . . . . . . . . . . . . . . 8.3.2. Sedimentation coefficient . . . . . . . . . . . . . . . . . . . .
243 244 245 245 246 241 248 249 250 250 251
Chapter 9 . Future prospects for density gradient centrifugation . 254 9.1. 9.2. 9.3. 9.4.
Centrifuge design . . . . . . . . . . . . . . . Developments in centrifuge rotors . . . . . . . Developments in ancillary systems . . . . . . Uses of centrifugal methods . . . . . . . . . .
. . . .
. . . .
. . . .
. . . .
. . . . . . 255 . . . . . . . 257 . . . . . . . 259 . . . . . . 260
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . .
262
Appendix I . Manufacturers and suppliers of centrifuges. ancillary equipment and special chemicals . . . . . . . . . . 263 Appendix II . Glossary of terms used in density gradient centrifugation . . . . . . . . . . . . . . . . . . . . . 265 Appendix III . Density and sedimentation coefficients of rut liver cell organelles . . . . . . . . . . . . . . . . . . 269 Appendix IV . Theory of preparation of density gradients . . . . 271 References
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
274
Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . .
287
CHAPTER 1
Introduction to zonal centrifugation
1 .l. Thefirst applications of centrifugation in biology A pleasing surprise given to writers of this type of article is the discovery of useless, but to them novel, facts, like the use of centrifuges for separating biological structures by Miescher in 1872. Little attention was paid at that time to the structure of cells, and for many years the use of centrifuges was restricted to applications such as the separation of milk, the collection of precipitates and (from about 1930 on) the separation of large particles such as nuclei. At that latter period most biochemists directed their attention to the separation of purified fractions - especially enzymes - rather than to analysis of the structure of the cell. Accordingly, sophisticated analytical ultracentrifugeswere developed for testing the homogeneity of purified fractions, but preparative centrifugeswere chiefly used for the collection of precipitates. Two major advances paved the way for a more wide-spread use of centrifugal techniques. Firstly, the development of alloys with a high strength in relation to their density permitted centrifugation at high speed of much larger quantities of material. Secondly, the development of methods for examining biological specimens under the electron microscope (see Palade 1971) revealed the complexity of the internal structure of cells. Earlier workers had attempted to purify nuclei (Behrens 1932) and mitochondria (Bensley and Hoerr 1934) but light microscopy was the only method of assessing these preparations. Given the complexity of the structure of the cell and the essentially arbitrary nature of the fractions which were being 8
Ch. 1
INTRODUCTION TO ZONAL CENTRIFUGATION
9
separated, what was needed was not ‘pure’ preparations, but a systematic and quantitative study of the distribution of subcellular particles and of enzymes between different fractions. De Duve (1971) has stated his belief that the insistence of workers such as Claude and Schneider and Hogeboom (1951) on the necessity for quantitative experiments ensured that the better understanding of the internal structure of the cell was rapiddly followed by an understanding of the biochemical role of the component parts. These early studies were all performed using the technique which was later called differential centrifugation or, more properly, differential pelleting (Reid 1972b). This method was refined by De Duve and his colleagues, and the ‘mitochondrial’ fraction was resolved into a heavy fraction containing mainly mitochondria and a ‘light’ fraction enriched in lysosomes (de Duve and Berthet 1954; de Duve et al. 1955). The importance of these advances is seen in the vast volume of work on the fractionation of different types of cell which followed. While differential pelleting has been an enormously useful technique for cell fractionation, a number of workers realised that only particles differing considerably in size could be separated in this way. Alternative techniques, rate and isopycnic zonal centrifugation were, in fact, proposed at about the same time as the differential pelleting scheme was developed, but the application of these methods was limited by the apparatus available (Anderson 1956; Allfrey 1959; de Duve et al. 1959). The full potentiality of density gradient centrifugation only began to be realised with the development of high capacity swing-out rotors and zonal rotors during the early 1960’s.
I .2. Centrifugal techniques 1.2.1. Analytical ultracentrifugation As we have mentioned, early high speed centrifuges were mainly used for the study of ‘exthcts’ from cells. Both this use and metallurgical limitations meant ‘that the volume of specimen had to be minimised. Rotors were designed with transparent windows so that the distriSirhiri I iiidc,x p 297
10
DENSITY GRADIENT CENTRIFUGATION
bution of particles in the centrifugal cell could be examined during centrifugation. A typical example of such an analytical rotor is shown in Fig. 1.1. The cells in such a rotor are filled with a uniform suspension of the mixture to be analysed. The rotor is
Fig. 1 . 1 . (left) A rotor for an analytical ultracentrifuge. (right) A high-speed preparative centrifuge rotor (Beckman type 65) is shown for comparison.
accelerated to its operating speed. Particles move at a rate determined by their size and shape and by the centrifugal force. Thus, if the cell was initially filled with a uniform suspension (Fig. 1.2A1) containing only one type of particle, a clear zone will appear to move slowly down the cell as the particles that were in that region sediment away (Fig. 1.2B1). If a mixture of particles were initially present, each type of particle will sediment at its own speed, so that after centrifugation the distribution will be as shown in Fig. 1.2B2. If some generalised property such as refractive index or ultraviolet absorbtion is measured, then patterns similar to those shown in Fig. 1.2C will be obtained. If an optical system sensitive to changes of refractive index is used (Schlieren optics) then the output will show a series of peaks (Fig. 1.2D). The latter is the form of output normally chosen, but it is important to realize that these peaks do not represent a zone of particles moving down the cell through a clear supporting medium, but the ‘back end’ of a sedimenting block of
Ch. 1 la1
1
11
INTRODUCTION TO ZONAL CENTRIFUGATION lbl
IC)
.:*.. . ..-.. .’.*..
2m Centrifuaal field
I
Fig. 1.2. Separations in an analytical rotor. The particles are initially distributed uniformly through the cell (A). As centrifugation proceeds, the particles sediment down the cell. Each type of particle will sediment at a distinctive rate. Thus as each group of particles sediment, a series of interfaces will form (B). These interfaces will appear to sediment through the cell at the same speed as the particles with which they are associated. The interfaces may be detected either by measuring the distribution of particles through the cell using some property common to all the particles such as ultraviolet absorbance (C) or by using an optical system (Schlieren optics) which provides an output which is related to the rate of change of refractive index (D). The latter system is less sensitive, but the output is more easily interpreted.
particles. As will beseen later, the use of an initially uniform suspension and the measurement of the clearance from the cell circumvents many problems. Analytical ultracentrifugation has been greatly developed from this essentially simple basis, but not in such a way as to fall within the scope of this article. Readers who are interested will find more extended accounts in books and articles by Schachman (1959), Trautman (1964) and Bowen (1970). 1.2.2. Differentialpelleting Differential pelleting is similar in principle to separations in an analytical ultracentrifuge. The centrifuge tube is filled with a uniform suspension.During centrifugation particles move down the centrifuge tube and pellet on the bottom. Ideally, centrifugation is continued S ~ h Ipr,rd
2f
00 VI
86
DENSITY GRADIENT CENTRIFUGATION
been used for the isopycnic banding of chromatin (Hossainy et al. 1973). Its effect on other subcellular structures is not known. 3.3.3. High molecular-weight organic compounds One solute stands out above all others. Ficoll (Pharmacia Ltd., Uppsala, Sweden) is a co-polymer of sucrose and epichlorhydrin with an average molecular weight of about 400,000 and was especially developed for density gradient centrifugation. It has little effect on biological particles and, unlike sucrose, does not inhibit enzymes even when it is present in high concentrations (Hartman et al. 1974). Ficoll, as purchased, is contaminated by some low molecular weight material which should be removed by dialysis before use. The viscosity of Ficoll solutions is higher than that of sucrose solutions of the same density, but this is counterbalanced by the fact that Ficoll does not penetrate biological membranes and only exerts a tiny osmotic pressure, so that membrane-bound particles band at lower densities in Ficoll than in sucrose. However, the tonicity of Ficoll rises exponentially with concentration, so that the tonicity of concentrated solutions is very high (Bach and Brashler 1970). With rate-zonal separations, the use of high molecular-weight solutes such as Ficoll allows a greater concentration of material in the sample zone than can be maintained with a solute such as sucrose as ‘sedimentation in droplets’ does not occur (see 0 2.1.2). Glycogen gradients have been used for the isopycnic banding of mitochondria and lysosomes (Beaufay et al. 1964). Glycogen gradients are difficult to handle as glycogen sediments markedly at quitea low centrifugal force and may be considered obsolete. Dextran gradients have been used for the isopycnic banding of microsomes (Graham 1972, 1973). The separations achieved are equivalent to those obtained in other tissues with Ficoll gradients. As the cost of dextrans and of Ficoll is about the same, there would seem little to choose between the two materials. WhileFicoll haslittleeffecton the functionof isolated cell organelles, it may be damaging to some living cells (Mathias et al. 1969; see also Ch. 7). A number of materials have been investigated for sepa-
Ch. 3
CONDITIONS FOR SEPARATIONS
87
rating such sensitive cells, including colloidal silica and iodinated aromatic compounds, both of which are considered below. Bovine serum albumin is perhaps the safest of all, but solutions of sufficient density are difficult to prepare and, when formed, are very viscous. The authors have no personal experience of albumin gradients, the strong solutions required for isopycnic banding are, said to be, difficult to prepare (Mateyko and Kopac 1963). However, albumin can be very useful for rate sedimentation of living cells (Ch. 7). 3.3.4. Other types ojdensity gradient solute While most materials suitable for use as density gradient solutes can be fitted into the three categories listed above, a number of other interesting materials have been proposed (Table 3.1). A finely divided form of colloidal silica (Ludox, duPont, Wilmington, Delaware) has been used for the banding of whole cells (Wolff and Pertoft 1972a), mitochondria (Lagercrantz and Pertoft 1972) and lysosomes (Wolff and Pertoft 1972b) and several other purposes. This material has a number of advantages over other gradient materials. Ludox solutions have no appreciable osmotic pressure and do not penetrate biological membranes. Unlike Ficoll solutions, Ludox solutions are very mobile, so that much shorter centrifugation times are needed. The size of Ludox particles is sufficiently large for gradients to be ‘self-forming’ (4 2.2.2) at moderate centrifuge speeds. Probably for the latter reason the angle-rotor will give very satisfactory separations (Wolff and Pertoft 1972b). Unfortunately, Ludox has two disadvantages. Firstly, it cannot be used in centrifugal fields of more than about 100,000 g as the silica particles will pellet. Secondly, concentrated Ludox solutions are not stable between pH 4 and 7.5 which is the region of greatest stability for most biological structures. The properties of Ludox as a density gradient material are reviewed by Pertoft and Laurent (1969). Iodinated aromatic compounds, such as are used as X-ray contrast media, have also been used as density gradient media. The most commonly used contrast media are mixtures of the sodium and methylglucamine salts of three derivatives of triiodobenzoic acid, I,,/>,‘,I ,,,d,u-------------------1
Fig. 5.21. A versatile and inexpensive home-made gradient maker. The major component is a pump with two separate and adjustable channels. (From Hinton and Dobrota 1969.)
is a very simple exponential gradient former which was first described by Anderson & Rutenberg (1967). It needs a supplementary pump. The second is a more elaborate instrument with a built-in pulse-free piston pump and a magnetic stirrer. It can produce gradients which are either linear or exponential or a combination of the two. The pump can also be used to load the sample by reversing the flow and withdrawing the cushion out (at a reduced flowrate) while the centre line is dipped in the sample which is sucked into the rotor. Although extremely well made and designed, its price is such that it has to compete with the more versatile fully variable gradient formers. Sorvall GF-2. In specification is very similar to the complex IEC instrument. Although cheap it does need a separate pump. Buchler Zonal Varigrad 2-5 18 1. Suitable for linear and exponential gradients. Suh/<wr nrdc., y 287
182
DENSITY GRADIENT CENTRIFUGATION
3) Complex, commercially-availableinstruments which can form any gradient profile. These are naturally more expensive than .the previous group but are the most versatile. Beckman model 141. The shape of the required gradient is cut out on a sheet of phosphor bronze which then controls the piston pump stroke and therefore the relative amounts of light and heavy solutions pumped. In spite of being one of the oldest available, the precise metering of the piston pumps makes it still one of the most suitable for zonal work. MSE, model Z-100A. This also requires the exact shape of the gradient to be cut on a sheet metal template, which then controls the relative flow-rates of two pumps. A slight snag with the instruments lies in the use of peristaltic pumps whose flow-rates require to be calibrated rather frequently. LKB Ultrograd. Represents a novel approach to gradient making. It meters the relative quantities of light and heavy solutions simply by the length of time a solenoid valve remains open to either solution. This valve is in turn controlled by a ‘graph reader’ on which the gradient is cut in black paper. A pump (which is a separate item) then draws either light or heavy solution, mixes them and pumps the resulting gradient into the rotor. Problems could well arise due to the pump having to suck the viscous heavy solution through the valve and also in the efficiency of the mixing. Isco Dialagrad. The only instrument which does not require the profile to be plotted on a graph or a metal template. The gradient is ‘dialled’ in ten points along a percentage scale, 0 being the light and 100 being the heavy solution. Two pumps then meter the appropriate amounts of the two solutions. Between the preset points the gradient is electronically smoothed to give a curve rather than a series of steps. 5.1.4.2. Pumps The pumps needed for zonal work fall into two
categories: those suitable for pumping gradients and those for circulating coolant round the rotating seal or other equipment.
Ch. 5
ZONAL ROTORS
183
The requirements for a pump for gradients are as follows: it should be able to handle liquids of different viscosities, it must cope with backpressures as high as 35 psi (2 atm.), the flow-rate should be adjustable in the range of (r60 ml/min and the hold-up volume of the pump must be as small as possible. Most piston and diaphragm pumps satisfy these criteria, but the pulsing flow may in rare cases cause leakage at the seal. The pulsing can be reduced by either a pulse suppressor or, simpler still, by using a piece of soft silicone rubber in the line between the pump and the rotor. Peristaltic pumps give a more even flow but are much more affected by variations in back pressure and viscosity of the liquid being pumped. Of the piston pumps available, we have" found the Micro Pump series 2 (Metering Pumps Ltd.) most suitable. It offers positive displacement (up to 250 psi.), adjustable, reasonably accurate flow-rates and most important, the facility to mount up to six pump-heads on to one drive, motor. We have described the application of such a pump, fitted with two pump heads (see Fig. 5.21) for a simple home-made gradient maker (Hinton and Dobrota 1969). To ensure that accurate flow-rates are maintained with this and other piston pumps, the outlet must at all times be positively pressurized. If this is not done, the solution can syphon through the pump since the zonal rotor is often at a lower level than the gradient reservoir on the bench. To prevent this, either a backpressure valve should be fitted in the outlet line or, place the outlet higher than the inlet. The ISCO Model 300 diaphragm pump should also be suitable. Although peristaltic pumps cannot deliver accurate flow-rates or pump against a high backpressure, they can nevertheless be quite good for zonal work. The most popular peristaltic pump for this application seems to be the Hiloflow (Metering Pumps Ltd.). Cheap centrifugal pumps are adequate for circulating coolant round the rotating seal and other ancillary equipment. A good example of such a pump is the Grants model P2. 5.1.4.3. Refractometers and density meters One reasonably good flow-through refractometer is no longer available. This was the Hilger Siihiwr bidex p. 287
184
DENSITY GRADIENT CENTRIFUGATION
& Watts model 550 which has been discontinued. Our own instrument has proved quite rugged over the last seven years of continual use (and misuse). The major problem with the instruments available today stems from the fact that they were primarily designed for use as detectors in chromatography. Consequently, sensitivity is extremely high (down to 16 x 10-8 RI) and the very narrow connections and the tiny flow-cell(50 pl or less) cannot take the high flowrates normally used in zonal work. Therefore, a difficult modification has to be made to reduce the sensitivity and only a part of the gradient to be monitored can be passed through the cell. The latter means that a stream splitter and a second (accurate) pump may be needed to sample only a part of the gradient. This portion of the gradient must either be fed back into the main line or be wasted, in which case its volume has to be known. A range of these instruments is available from Waters Ass., Atago Ins., Laboratory Data Control, Perkin-Elmer, Phoenix Precision Inst. Co., Varian, Winopal, etc. The Winopal refractometer has been used successfully for zonal work by Bachofen (1974). A number of density meters are available but they are all large industrial instruments not readily adaptable for small laboratory use. We will therefore not list them here. The instrument with the nearest specification to ideal seems to be the Anton-Paar DMA-10 density meter. However it has no recorder output but only a digital display suitable for single readings (although a whole zonal gradient can be pumped through the measuring cell). We believe that a new instrument with an anologue output, is now available but at present have no details. Such instruments measure density and are not affected by changes of gradients solute. The /3-ray absorption density meters described by Atherton et al. and Cope and Matthews (1973) are both affected by different solutes and therefore need to be calibrated for every class of solute. In practice this problem of changing scales is not serious as most zonal workers routinely use only one gradient material. Of the two above instruments the Cope & Matthews one appears simpler and therefore cheaper although neither are commer-
Ch. 5
ZONAL ROTORS
185
cially available. However, we believe that the Atherton instrument can be made to special order. Some while ago the idea of weighing a loop of tubing with a strain gauge whose voltage response was linear with density was investigated (see Dobrota and Reid 1971) but nothing has appeared as hardware. To sum up, although none of the instruments seem ideal we hope that the above information will prove useful for the interested user. 5.1.4.4. Spectrophotometers,colorimeters and UVmeters. The range of instruments suitable for locating the separated bands is so large that we could not possibly list them all. As a general point they must have provision for a good flow-through cell, if possible have a reasonable range of wavelengths and must have a recorder output, and also should be compact enough to fit onto a trolley. Manufacturers include Pye Unicam, Perkin-Elmer, Carey-Varian, Beckman, Gilford, Cecil, etc. We particularly like the compact Cecil 404 model and some of the Beckman instruments. The more sophisticated, and more expensive, instruments such as the Gilford can cope with very high absorbance readings. The recorder output of most colorimeters and some spectrophotometers will be linear with transmission and may therefore need converting to absorbance. This can be done with a suitable logarithmic converter. Chromatography UV monitors like the LKB Uvicord, Gilson, and Isco may also be used. While the Gilson, Isco (which reads in absorbance) and Uvicord I represent good value for money, the new Uvicord 111(although suitable) is as expensive as a spectrophotometer. Its purchase solely for monitoring zonal gradients could hardly be justified.
5.1.4.5. Flow-through cells. Beckman-RIIC market two flowthrough cells. Type BTF 5 has a variable light-path while the ‘multi-path flow cell’ has four fixed light-paths (see Fig. 5.22). The latter cell can only be fitted in a Beckman spectrophotometer (although the outside dimensions are the Same as a standard 10 mm cuvette). However, a cheap non-adjustable flow-through cell, such as Sehiecr nrdi,r p 287
186
DENSITY GRADIENT CENTRIFUGATION
Fig. 5.22. The Multipath flow cell. (By permission of Beckman-RIIC.)
the Hellma type 154, with a 10 mm lightpath, a hold-up volume of about 1 ml and fitting a standard cuvette holder is ideally suited for zonal work. Similar cells are also available from Scientific Supplies. 5.1.4.6. Thermometers. For measuring the in-line temperature of
gradients suitable thermocouple electrical thermometers can be obtained from Gallenkamp (Cat. No. TK-075)and Electroplan (Comark range). 5.1.4.7. Perfusor pumps. The Sage perfusor pumps (sold by A.R. Horwell) and the Braun perfusor (Shandon-Southern) can be useful for pulse-free injection of precious zonal samples and also for unloading swing-out tube gradients. 5.1.4.8. Stroboscopic lights. These can be extremely useful parti-
cularly with the transparent rotors. Firstly they are ideal for revealing particle aggregation if it should occur while to the eye the bands might have appeared as crisp well defined bands. Secondly they can be used to check the rev/min of any spinning rotor. Suppliers of strobes, amongst many include Dawe Instruments, Electronic Applications Ltd., Electroplan, General Radio, etc. 5.1.4.9. Recorders. In view of the need to monitor two parameters, a two channel (or even more) recorder is most suitable. The two pens
Ch. 5
Z O N A L ROTORS
187
should provide the trace on the same chart paper, and there must also be an event marker pen linked to the fraction collector. Good examples of such recorders are the Rikadenki Model 241 (TEM, Sales) and the two-channel Smiths Servoscribe. Preferably the recorder should have a fairly wide range of input, i.e. 5 mV to 5V (DC), and a reasonable adjustment on both zero and full span so that odd (e.g. 85 mV) inputs can be matched up to full-scale deflection on the recorder chart. The speed of pen response does not need to be fast since both gradient and absorbance traces are gentle curves. 5.1.4.10. Coolingequipment. Although a supply of ice should suffice, a permanent supply of ice-cold-water from a bath may be advantageous. A suitable refrigeration unit is available from Grants, model CC-15, and can supply enough cold water (with a suitable circulating pump) for the seal, gradient cooling coil and the fraction collector jacket. Most high-speed centrifuges need a supply of cooling water for the diffusion pump condenser and various bearings. For example a Super Speed 65 or 75 (MSE) requries water (2l/min) at not less than 20 psi (1.5 atm.). With some Beckman machines there is a maximum pressure which must not be exceeded otherwise a flood will result from a burst diaphragm in the pressure switch. If the continuous use of a cold water supply is too wasteful (in practise the tap tends to be left on all the time), a simple recirculating system can be constructed. This would need a tank, a car-type radiator, tubing and an electriccentrifugal pump (as in a washing machine) to circulate the water from the tank to centrifuge then to radiator and back to the tank.
5.1.4.1I . Minor accessories. Tubing suitable for zonal flow lines should be not less than about 3 mm bore since it could cause excessive pressure build up. Cheap PVC tubing such as No. 6H (Portex Ltd.) can be used for most of the lines but since it becomes rather rigid when cold, soft silicone rubber tubing (Bsco-Rubber Ltd.) is preferable for the connections to the feed-head. Simple plastic Siihpw inct~~x p. 287
188
DENSITY GRADIENT CENTRIFUGATION
connectors for joining up lengths of tubing can be bought from most laboratory suppliers. Tubing clamps, for securing tubing connections are obtainable from Uniclip Ltd. Plastic threeway taps, which are most useful for the two feed lines to the feed-head since they allow reversal of flow and easy connection of syringes, are supplied by Henley Medical Supplies Ltd. and Portex Ltd. (through agents). Artery forceps (also called Spencer-Wells,and obtainable from any surgical instrument supplier), are invaluable for rapid and effective clamping of tubing in case of a blown connection or ather emergencies. The gradient cooling coil can be made from stainless-steel tubing of approx. 3 mm bore. This can be supplied by the Oxford Instrument Co. 5.1.4.12 Fraction collectors. As with spectrophotometers, the range
of fraction collectors is wide and well known to biochemists who often have personal preferences. Apart from stating that compactness, cooling facility, a capacity for large tubes (up to 50 ml.) and the provision of an event marker are all important we leave the choice to the reader. The major problem of collecting fractions of equal volume (see 85.1.3.6.0 can be solved by using an Isco Volumeter. This device detects the level of the liquid in a burette-like vessel and then opens a solenoid-operated tap allowing the liquid to drain into the fraction collector tube. A useful modification to allow very rapid drainage into the fraction collector tube is to increase the bore of the Teflon delivery tap. If this is not done the volumes of the fractions can get progressively bigger, with the increase in gradient concentration. An alternative approach is to use a fraction collector of the Gilson Escargot type which detects the volume in the collector tube which is a molded uniform plastic tube. In view of the relatively high flow rates used in zonal work, some spillage may occur between tubes as the fraction collector changes.
Ch. 5
ZONAL ROTORS
189
Although this problem does not arise with the Volumeter, with other systems it may be serious enough to warrant the instalation of a flow-arrest device, such as that available for the LKB Ultrograd. 5.1.4.13. Integrator This accessory, which is electrically connected
to the centrifuge, calculates and displays the time intergral of the square of rotor velocity (w’t). Although expensive, it can be useful for exacting analytical work, especially since it can be preset to switch the centrifuge off at any selected value. It is available for the MSE. Beckman and Christ machines.
5.2 Reorienting zonal rotors When the tubes of a swing-out rotor move from a vertical to a horizontal position as the rotor accelerates, they mechanically move the tube contents into the new position dictated by the centrifugal field (see also 81.3.3). Thus the whole tube is reoriented, leaving the contents undisturbed. However, if a gradient is loaded into a zonal rotor at rest and the rotor is then accelerated, the gradient will in fact reorient from a vertical to a horizontal position (see Fig. 5.23). This will occur in any system, including angle tubes, and can be successful provided that the rotor shape is correct and accurately controlled rates of acceleration and deceleration can be achieved. Anderson et al. (1964) and Fisher, Cline and Anderson (1964) were among the first to report the successful reorientation of density gradients when they described the use of angle rotors for isopycnic banding of DNA and phage on CsCl gradients. Later, the first reorienting zonal rotor was described by Elrod, Patrick and Anderson (1969). Gradient profiles were checked before and after the spin and found to be almost identical, thus confirming that gradients in zonal rotors would, under the correct conditions, reorient successfully. The merits of this technique are that if the rotor can be loaded and unloaded successfully, no rotating seals, such as the dynamic Rulon seals, are needed. Also, loading and unloading need not be done in the centrifuge, they could be done on the bench or in the Suhlc~irfriCr p 287
190
DENSITY GRADIENT CENTRIFUGATION
Ch. 5
ZONAL ROTORS
191
Fig. 5.23. Sequence of events in operating an RK rotor, showing reorientation and continuous-flow banding. A density gradient is loaded into the rotor at rest (A). The gradient reorients vertically as the rotor is carefully accelerated, up to its operating speed (B). A sample is pumped in at the top. As it flows over the top of the gradient, particles sediment into the gradient and are ‘captured’. while the eMuent Suhjrcr index p
2x7
192
DENSITY G R A D I E N T CENTRIFUGATION
flows out through the bottom seal (C). Sample flow is stopped and the captured particles are allowed to equilibrate to their isopycnic point (D). The rotor is decelerated (E) and the gradient reorients to its original horizontal position (F). This process does not disturb the particle bands. The rotor is unloaded, at rest, by applying air or water pressure to the top of the rotor and using a pump to control the flow ( G ) .
cold room. Therefore the whole technique of zonal centrifugation could be enormously simplified. Anderson (1966a) reported that a small reorienting rotor of 100 ml capacity had been tested at 141,000 revs/min. Most of the problems are purely technical. Firstly a smooth and controlled rate of acceleration and deceleration is needed, especially in the low speed range, i.e. up to 500 or 1000 revs/min. This is most easily achieved when the centrifuge is driven by an air or oil turbine, as in the K, RK series of reorienting continuous-flow rotors. With electric-drive centrifuges we have encountered a slight problem (see below). Since the rotor is loaded and unloaded at rest and thus the gradient is not stabilised by centrifugal force as in the classical zonals, the channels for loading and unloading must be uniform and of precisely the same diameter. A further, but only theoretical objection, is the greater likelihood of droplet formation at rest before a sufficient g is reached to prevent it (see $2.1.2). Reorientation in zonal rotors is already in common use. Rather remarkably the rotors used (K, RK and J) are tall cylinders, which would appear to be quite unsuitable, and yet according to many workers notably Cline, there is apparently little loss in resolution. It is therefore evident that the technique can be made to work, but its usefulness for a wide range of separation problems (especially rate sedimentation) remains to be demonstrated. Reorienting rotors: Sorval SZ-14. This rotor was developed by Dr. P. Sheeler and has been commercially available for some time now. It can be operated in a number of Sorvall centrifuges, including the RC2-B in which it has the maximum operational speed of 19,500 (maximum g 40,500). The total capacity is 1,400 ml and the usable centrifugal path is 5.3 cm. The minimum radius of the rotor is 4.2 cm meaning that
Ch. 5
ZONAL ROTORS
193
a reasonably high g is available in the sample region thus making the rotor suitable for isopycnic work. We have no personal experience of operating this rotor, but have no reason to suspect any major difficulties. The rotor shape is shown in Fig. 5.24.
Fig. 5.24. Diagramofthe SZ-14 Sorvall reorientingrotor.Upper Fig. shows the sample application and, on the right half of the rotor, the separated bands. Lower Fig. shows the rotor at rest after reorientation ready to be unloaded.
The maximum g attainable would put this rotor in terms of its applicability somewhere between the slow speed A, HS and the B-XV zonals. Unfortunately it is still difficult to assess its absolute usefulness mainly because relatively few publications have quoted it, and of these the majority seems to originate from the team which designed it. They have demonstrated (Sheeler and Wells, 1971; Wells, Sheeler and Gross, 1972) that it can be used for separating mitochondria, lysosomes, nuclei, liver glycogen and even rough and smooth membrane vesicles. The one worry which we would have about this rotor is purely theoretical and not founded on any practical experience. This concerns Sirhirrr irrde.~p. 287
194
DENSITY GRADIENT CENTRIFUGATION
Fig. 5.25. Reorientation of sucrose gradients in a B-XV rotor. Graphs 1-5 show the profiles of step gradients, 300 ml water, 400 ml 0.5 M sucrose, 400 ml 1.5 M sucrose, approx. 560 ml of 2 M sucrose, monitored during loading (----) and during unloading -( ). In each case the gradient was left at about 3,000 rpm for 15 min.
Ch. 5
195
ZONAL ROTORS
the method of unloading. The rotor has a core and six septa which are provided with channels leading to the tapered bottom of the rotor. Having experienced sectorial unloading with an A-XI1 rotor (see Fig. 5.19) at approximately 500 g, we suspect that under 1 g there could be differential unloading through these six channels, thus resulting in a loss of resolution. K , RK series. These rotors have already been described in $5.1.2.8. They are all used as dynamiczonals or as true reorienting rotors. While the large K and RK are better suited for continuous-flow work, the J rotor can be applied in very much the same way as a B-XIV zonal batch rotor (Cline 1972, personal communication). B-XZV and XV as reorienting rotors. Beckman and MSE rotors can be supplied with diagonal septa channels for reorientation work. Klucis and Lett (1970) have used a B-XXV zonal rotor (a special version of a B-XV designed specifically for DNA work) in a reorienting mode; although they do not discuss the reorientation it must have been successful. In view of the scanty information available, we decided to test the feasibility of this method, using an aluminium B-XV, running in a SS65 (MSE) centrifuge, which was fitted with a slow accelerate/decelerate control unit. (When using this control, the motor brushes must be in perfect working order, otherwise acceleration and deceleration will be uneven.) These tests consisted of simply loading and unloading complex step gradients under various conditions (summarized in Fig. 5.25), and continually monitoring the gradient profiles on a recording refractometer. between loading and unloading. 1) Rotor used in normal zonal manner, gradient loaded and unloaded with the rotor spinning. 2) Gradient was loaded with rotor spinning. Rotor was then decelerated to rest, left for 15 min and accelerated to unloading speed, and unloaded. 3) Exactly as ( I ) except that the 300 ml of water was not pumped in as part of the gradient but was injected to the centre as with a normal sample. 4) Gradient was loaded with the rotor stationary. It was then accelerated and after 15 min, unloaded while spinning. 5 ) Gradient was loaded with rotor spinning. After deceleration it was unloaded at rest, by pumpingcushion to the bottom anddisplacingthegradient upwards. 6 )Alineargradientwasloadedwiththe rotor at rest. After acceleration it was unloaded normally with the rotor spinning. .Svhii,ir
in&.r
p. 287
196
DENSITY GRADIENT CENTRIFUGATION
From the results two distinct conclusions can be drawn. First, let us state categoricallythat gradients are reoriented successfully, indeed the shape of the gradient is in some cases preserved better than with dynamic loading. When a sharp step gradient is introduced dynamically and then unloaded, the first step becomes almost a smooth curve, while the step is much better defined if the identical gradient is loaded statically and then reoriented. The most likely reason for this is that when the layers of gradient enter the rotor, under dynamic loading, their surface areas at the wall periphery are so large (radial dilutions effect) that solute diffusion destroys the step. The surface area of the bottom of rotor - the chamber - is much smaller than that of the wall thus explaining the better preservation of profile in the case of static loading. However, the second conclusion is that there will always be some loss of resolution when a B-XIV or B-XV rotor is unloaded at rest. The reason is that the top of the rotor chamber is completely flat, and as the discrete layers of gradient are pumped up (or down), they are held up at the corners and mixed in with the successive layers. Were the top of the rotor slightly tapered much like a tube-unloading device, unloading could be without loss of resolution. In emergencies, when the rotor has stopped and cannot be restarted, smooth gradients may be recovered without disastrous mixing if displaced at a very slow rate (less than 10 ml/min).
CHAPTER 6
Assay of fractions separated by density gradient centrifugation
6.1. Enzyme and chemical assays on fractions An article of this size cannot include all the methods used in the assay of fractions separated by density gradient centrifugation. To assess a centrifugal separation, it is often necessary to assay markers for the various subcellular structures present in samples. We have discussed the choice of such markers elsewhere (Reid 1972; Hinton and Reid 1975). There are two general points which can usefully be made. Firstly, with the number of fractions normally obtained from each separation (20-60) it is hardly worthwhile to set up a separate AutoAnalyser channel for each assay. When chemical constituents such as protein are to be assayed, the fractions can be stored until material from several separations has accumulated, and it does then save time to set up an Auto-Analyser manifold. In some cases, such as enzymes releasing p-nitrophenol, a single AutoAnalyser manifold may be used for the assay of several enzymes. It is, however, tedious to have to perform several assays on up to 60 fractions without any analytical aids. Unstable components, especially enzymes such as glucose-6-phosphatase, are best assayed manually but using dispensors and repeating syringes as much as possible. However, care must be taken in using sampler diluters as some do not give reproducible results when used with fractions containing appreciable amounts of sucrose (Reid 1972~). We have discussed the Auto-Analyser because this is the only analyser which we ourselves have used. It would be interesting to know how useful modern discrete analysers (see Roodyn 1971) such 197
Siihjeil eiCr 1’.
287
198
DENSITY GRADIENT CENTRIFUGATION
as the Vickers D-300 would be for this application. While no single group could justify the purchase of such a machine, their speed and the ease with which they may be switched from one assay to another might make then suitable as a joint purchase by several research groups. Another analyser which would be of interest to a centrifugation laboratory, if only for its mode of operation, is the GeMSAEC analyser (Anderson 1969b). In this machine the reactions take place in a centrifuge rotor and are monitored continuously through glass windows inset in the rotor. The problem with this analyser is that the high density of some fractions separated from density gradients causes difficulty in mixing even at 1 g, and one would expect very severe problems to arise if, as in the GeMSAEC, one tried to mix sample and reagents during centrifugation. Another important point is that inert density gradient solutes such as sucrose may in high concentrations, interfere severely with the assay of enzymes and other tissue constituents. We consider such interference in more detail in Ch. 8.
6.2. Electron microscopic examination of fractions There is a tendency to underrate the usefulness of the light microscope in the examination of subcellular fractions. In fact, with particles the size of mitochondria or larger, light microscopy, especially with phase contrast, gives more information on the composition of fractions than electron microscopy. This is because one can rapidly inspect a representative sample of a fraction using a light microscope and recognise intact particles more easily than in the thin sections used in electron microscopy. However, particles smaller than mitochondria can only be resolved by the electron microscope. The great problem with electron microscopy of fractions separated by density gradient centrifugation is to obtain a representative sample. When the particles are pelleted from the separated fractions, the pellet will naturally tend to be non-uniform with the largest particles at the bottom of the tube. When the pellet is broken up during fixation and embedding, the orientation of the fragments will be lost.
Ch. 6
ASSAY OF SEPARATED FRACTIONS
199
In addition, fragments of the pellet will tend to lie flat at the base of the capsule and consequently the tendency will be to cut across a pellet rather than through its thickness. Any particular section may, therefore, appear homogeneous, in spite of considerable striation in the original pellet. There are a number of ways of minimising or circumventing this problem. When fractions are collected by pelleting one should, 1) dilute to give the thinnest possible pellets, 2) break up the pellet as little as possible during fixation and embedding, 3) cut across the thickness of the flakes, and 4) cut sections at several different depths in the block. Alternatively, one may avoid pelleting and separate the particles from the medium by filtration rather than by centrifugation (Baudhuin et al. 1967). This gives more even sampling than pelleting and makes it simpler to maintain the orientation of the specimen during fixation and embedding. Finally, one may abandon sectioning altogether and use negative staining methods (Whittaker et al. 1964). The disadvantage here is that the material must be examined immediately after separation and the images are more difficult to interpret than those obtained after thin sectioning. A second problem in electron microscopic examination of subcellular fractions is to identify the particles. For example, while it is simple to recognise intact mitochondria, damaged or fragmented mitochondria may be difficult and even impossible to recognise. The answer lies in specific cytochemical staining methods for enzyme activity, but there are many problems associated with their use although excellent results have been obtained by Leskes et al. (1971) and El-Aaser et al. (1973). Poor penetration of the substrate into jelly-like pellets can be avoided by resuspending the particles, doing the cytochemical reaction in free solution and pelleting after fixation with glutaraldehyde. The pellets can be stained with osmic acid and embedded in the usual way. Because the particles are collected by centrifugation after the cytochemical reaction, one must again be aware of striation.
200
DENSITY GRADIENT CENTRIFUGATION
6.3. Assessment of resultsfrom density gradient separations The presentation of the resultsof analytical separations usually involves the study of the distribution of a particular cell component - often an enzyme - among the separated fractions. The patterns can then be compared with those of the established ‘markers’ (see Reid 1972; Hinton and Reid 1975). The most usual (and convenient) method of presenting results is to plot the total activity (or amount) per fraction against fraction number, but if the fractions are not of equal size the activity (or amount) per unit volume should be used. As well as the distribution of each component across the gradient, one also wishes to gauge the reliability of the experiment. We find two control values especially useful. Firstly comparison of the enzyme activity of the homogenate with the activity obtained with a similar tissue in earlier experiments serves as a control on systematic errors in the assay methods. One rapidly learns to distinguish the normal variation between animals from the truly ‘odd’ result which requires further investigation. The specific activity of the homogenate (i.e. the activity per mg. protein) is usually more constant than the activity per gm. wet weight as it is less subject to errors arising from poor homogenisation and from adhesion of liquid to the surface of the tissue. The second essential quality control figure is the recovery, the sum of the activities in the fractions divided by the activity in the material loaded onto the gradient. The latter is estimated from a sample, taken before separation, stored at the temperature used in centrifugation and then assayed in parallel with gradient fractions. If the recovery is much more or much less than loo%, then one must exercise great caution in interpreting the results until an explanation for the activation or loss in activity is discovered. While presentation of the results of analytical separations simply as activity per fraction is usually sufficient, sometimes other methods of plotting will give better results. After separation by isopycnic banding one may wish to correct the results for the effect of nonlinear gradients. This may be done simply by dividing the activity in any fraction by the density difference across that fraction and plotting
Ch. 6
ASSAY OF SEPARATED FRACTIONS
201
the resultant figure (activity/unit density difference) against density. The result of this sort of transformation can be seen in Fig. 7.7. However, it is usually better, in such a case, to repeat the separation on a linear gradient. The plotting of specific activity (activity/mg protein) or of purification (the specific activity/specific activity of the same enzyme in the homogenate) as used in presentation of the results from differential pelleting (de Duve 1966) is dangerous. Thus the experiment shown in Fig. 6.1 would appear to show 5’-nucleotidase
. c
‘E
Fraction no.
Fig. 6.1. An illustration of the dangers of using the specific activity of an enzyme as a guide to its distribution. A rat liver post-nuclear fraction was fractionated by centrifugation in an HS zonal rotor as described by Burge and Hinton (1971). When the specific activity of 5’-nucleotidase is plotted (A) a ‘peak’ is shown in the central (lysosome-rich) part of the gradient. The actual distribution (B) shows that a negligible proportion of the 5’-nucleotidase is actually recovered in the lysosome-rich region. Upper part : 5‘-nucleotidase specific activity. Lower part: 0__ 0, 5’-nucleotidase: . . . . . ,acid P-glycerophosphatase; . . . . . , protein. ~
.
Slrh,w/ b v k . ~ p . 287
202
DENSITY GRADIENT CENTRIFUGATION
in the lysosomes when the specific activity is plotted. In fact, further fractionation gives no evidence for this and in any case the amount of 5’-nucleotidase activity in the lysosome-rich region is tiny. One must remember that de Duve’s method for presenting the results of differential pelleting experiments was developed specifically for the localisation of lysosomal enzymes, and while very useful for this purpose can be very misleading outside its proper context. In our experience, the most valuable alternative method for plotting results from separation experiments is to plot the activity in each fraction as a percentage of the total activity recovered from the gradient. This has the effect of reducing all the distributions to a common scale and makes it easier to distinguish small, but significant differences in distribution. Thus in the experiment illustrated in Fig. 6.2 it would be difficult to be sure that acid ribonuclease was differentlydistributed from acid phosphatase, from the plot of activity, but plotting the results as a percentage of total activity makes the difference clear. For preparative separations, reporting requirements are that results be given for markers of potential contaminants as well as for the particles being separated, and that the results for each marker should be presented, a) as a percentage of the homogenate activity recovered in the final preparaion (and preferably at intermediate steps) and b) as a purification (i.e. the ratio of the specific activity in the final fraction to the specific activity in the homogenate). These figures will permit direct comparison with the results of other workers. Specific activities are meaningless in the absence of data on the homogenate.
6.4. Calculation of sedimentation coefficients Although ‘simple’ methods have been proposed for the calculation of the sedimentation coefficients of particles fractionated by rate zonal centrifugation (McEwan 1967; Halsall and Schumaker 1969; Young 1972; Funding and Steensgard 1973), the authors doubt their value. Fortunately there are a number of well developed programs
Ch. 6
203
ASSAY OF SEPARATED FRACTIONS
uMolssImin E260/min
10.4
Fig. 6.2. An illustration of the advantages of plotting all results on a common scale. A mitochondria1 + lysosomal fraction from the livers of rats injected 3 days before death with Triton WR-1339 was fractionated by centrifugation in an HS zonal rotor as described by Burge and Hinton (1971). Acid B-glycerophosphatase (A-A) is concentrated in lysosomes which sediment slightly faster than those which contain acid ribonuclease (0----0). This difference is shown more clearly when activities are plotted as a “6 of the activity recovered from the gradient (B) than when absolute activities are plotted (A).
for these calculations, and if a computer is available, one can usually find a program which will run on that particular machine. It is, however, frequently not neccessary to use a computer program. For reasons that will be discussed later (0 8.3.2) we feel Suhjrcr index p . 287
204
DENSITY GRADIENT CENTRIFUGATION
that one cannot, with present centrifuge equipment, use density gradient centrifugation for the primary determination of sedimentation coefficients, for this purpose one must use an analytical ultracentrifuge. What can be done is to compare sedimentation coefficients, so that if the value for one component of a mixture is known, one may calculate the coefficients of other components. If isokinetic or, with zonal rotors, isovolumetric gradients (see 5 3.4) are used, one may carry out this calculation by linear interpolation. With the linear 0 . 5 1 .O M (1 5-3073 gradient which is commonly used with swingout rotors, even the errors introduced by linear extrapolation are not very large, providing that the particles whose sedimentation coeffcient is known are not too dissimilar in size to the ‘unknown’ particle. Much greater problems arise with zonal rotors because of radial dilution as well as non-linear sedimentation, so that it is usually impossible to use linear interpolation. However, one of the authors having dabbled with the problem, suspects that ‘simple’ methods for calculating sedimentation coefficient are usually more fun for the inventors than useful to other workers. The choice is really between proper computing of the sedimentation coefficient and obtaining a rough estimate by comparing the radial distance moved by the ‘unknown’ particle with the distance moved by a particle of known sedimentation coefficient. As the most frequent use of the sedimentation coefficient is as a label for particles which can be identified in no other way, a rough estimate is often quite sufficient. An accurate estimate of the sedimentation coefficient is only required when constructing mathematical models of density gradient separations.
CHAPTER I
Applications of density gradient centrifugation
We shall not attempt to give a complete guide to all the problems to which the technique of density gradient centrifugation may be applied, much less to list all the approaches. Rather we will try to indicate the major areas in which gradient centrifugation may be useful and to discuss briefly the approaches which may be most helpful and the problems which may arise. Detailed information on separation methods is scattered through the literature, but the monograph on zonal rotors edited by Anderson (1966a), Volumes 1, 3 and 4 of ‘Methodological Developments in Biochemistry’ (Reid 1971, 1973 and 1974) and the book edited by Birnie (1972) form useful entries to the literature together with Chevrenka and Elrod (1972). In all cases the actual articles are biased very strongly towards the use of zonal rotors, but the introductions and discussions contain references to methods used with conventional rotors.
7.1. Separation of living cells In general, biochemists have tended to pay lip service to the biological evidence of the tissues of higher organisms, but have not been able to separate the different cell types in sufficient quantity and purity to enable any extensive biochemical studies to be carried out. There are two major steps in the separation of the different types of cell which make up a tissue. Firstly, the individual cells must be separated from each other by a procedure sufficiently gentle for the separated cells to retain their function. Secondly, a method must be found to separate the different types of cell from each other. The first topic 205
.Sirhii,ir
nid‘*.rp 287
206
DENSITY GRADIENT CENTRIFUGATION
is outside the scope of this article and it must suffice to say that the majority of published methods rely on the perfusion of the tissue with an enzyme such as collagenase (Munthe-Kaas and Seglen 1974; Crane and Miller 1974; Glick et al. 1974), pronase (Romrell et al. 1975) or trypsin (Aspray et al. 1975) which will break down intercellular connective tissue. A chelating agent such as sodium citrate or EDTA is sometimes included to facilitate disruption of the tight junctions. The most popular method for separating the released cells, and for the separation of free living and blood cells, is density gradient sedimentation either in a centrifugeor at unit gravity (Harwood 1974). Whole cells may be fractionated either by rate sedimentation or by isopycnic banding. It is hardly necessary to centrifuge when separating by rate sedimentation as cells sediment significantly at 1 g (Miller and Phillips 1969; Denman and Pelton 1973), Cells may also be fractionated by isopycnic banding, but one must take great care when using centrifugal fields greater then 15,000 g as cells may actually be pulled apart under their own weight (Mateyko and Kopac 1963). However, the greatest problem in the separation of living cells is the selection of the density gradient medium (see Table 3.1). As mentioned in 5 2.3, only density gradient solutes of a fairly high molecular weight are suitable for the separation of whole cells. Probably the most widely used compounds have been Ficoll and bovine serum albumin, their use is discussed in some detail by Harwood (1974). Both compounds are well suited for the formation of the supporting gradients used in rate sedimentation, for their high molecular weight will minimise instabilities in the sample zone (see 5 2.1.2). They are, however, less well suited for separations by isopycnic banding; the high viscosity of concentrated serum albumin solutions means that prolonged centrifugation is needed for the cells to reach their isopycnic banding densities, while the tonicity of Ficoll solutions rises sharply, and in a non-linear fashion, at high concentrations (Bach and Brashler 1970). Hence there has been some interest in novel density gradient media. Ludox (colloidal silica) has been used for cell separations (Pertoft and Laurent, 1969) but was found to damage spermatozoa, causing gelation and reducing motility (Bene-
Ch. 7
207
APPLICATIONS
dict et al. 1967). Dextrans have also been employed (Harwood, 1974) but may be expected to give the same problems as Ficoll. The most interesting group of compounds is the X-ray contrast media whose properties are described in +henext paragraph. The requirements for an X-ray contrast medium, low toxicity coupled with the capacity to form solutions of a high density, are very similar to the requirements for a density gradient solute. The commonest X-ray contrast media have all been used in density gradient centrifugation (see 9 3.3.4 & Table 3.1) although Metrizamide has been recently studied Litensively. The other media, notably Renografin (a mixture of sodium and methylglucamine diatrizoate), have been used for some time for the separation of microbial and blood cells (Hadden and Nester 1968. Tamir and Gilvarg 1966; see also refs in Hinton and Mullock 10%) Metrizamide would seem, on balance, to be the most suitable of the X-ray contrast media for use in density gradient centrifugation (Hinton and Mullock 1976), although none of them appear to cause serious damage to living cells. No further details will be given here, as separations in gradients of iodinated media have been covered extensively in articles in Rickwood (1976). All these compounds are expensive, so that financial considerations may prohibit their use in zonal rotors. To summarise, one may fractionate cells by rate o r isopycnic centrifugation. In the former case, bovine serum albumin or Ficoll are the media of choice. These media may also be used for isopycnic banding, but iodinated media such as Metrizamide are probably more suitable, as used by Munthe-Kaas and Seglen (1974) in the separation of liver cells (Fig. 7.1). Where neither simple rate separation nor isopycnic banding can separate the mixture, one can use the two techniques sequentially. Alternatively one could experiment with steep gradients approaching the banding densities of the cells as used by Boone et al. (1968); in this way it is possible to exploit differences both in density and in size to optimise separations between two populations of cells. In such cases computer simulation of the separation, as used by Boone et al. (1968), may be a great help in designing the gradient and deciding on the optimal centrifugation time. Sllh,'
t - I n
-
1.07
- 1.05 BOTTOM
FRACTION NO.
TOP
0
Y
Fig. 7.1. Distribution of rat liver cells in a Metrizamide density gradient. Freshlyprepared rat liver cells were incorporated into an 8 ml Metrizdmide gradient (density 1.05-1 .I6 g/cm’) and centrifuged for 20 min at 5000 revs/min at 4 in the tubes of a Beckman SW 40 rotor. A) Osmolarity; B) Density ( X----x ) ; parenchymal cells (0-0); non-parenchymal cells (0-0): Note 10-fold scale difference (from Munthe-Kaas and Seglen 1974).
7.2. Separation of cell organelles from mammalian tissues Because of the severe limitations on the amount of material which can be loaded onto a density gradient (see 9: 2.1.2), one does not normally attempt to separate an entire tissue homogenate by density gradient centrifugation. Rather the homogenate is first fractionated by differential pelleting. The fraction so obtained may then be further separated by rate o r isopycnic zonal centrifugation. The initial fractionation has the advantages both of simplifying the mixture and, in preparative experiments, of reducing the amount of contaminants so enabling more material to be processed in a single centrifugation.
Ch. I
APPLICATIONS
209
In spite of the comments in the previous paragraph one may sometimes wish to fractionate a whole homogenate, especially when dealing with a tissue in which the size of the various particles is not well established. An A-XI1 zonal rotor is probably the most suitable instrument for this purpose. The gradient we have tested is constructed in two parts, a fairly shallow initial section occupying about 2/3 of the diameter of the rotor followed by a steep final section (Hinton 1972). (Earlier descriptions of this gradient (Hinton et al. 1970 and 1971) contain errors.) Mitochondria, lysosomes and microsomes separate in the early part of the gradient on the basis of their sedimentation rate; in the latter part of the gradient, nuclei, fragments of connective tissue and red blood cells and plasma membrane sheets (which all sediment so fast that, if the separation were purely on the basis of sedimentation rate, would have pelleted on the wall of the rotor before the smaller organelles have left the starting zone) are separated on the basis of their isopycnic banding densities (Fig. 7.2). Such a separation may be useful as a first step in examining the distribution of some enzymic or chemical components among subcellular particles. If one is used to handling the A-XI1 rotor, the fractionation is much quicker than working through the scheme for differential pelleting and, except in the case of microsomes and cytosol, more information on the distribution of the unknown component is obtained. Generally, however, density gradient centrifugation is used for the further fractionation of organelle fractions separated by differential pelleting. 7.2.1. Subfractionation of a crude nuclear fraction and the separation of large sheets of a plasma membrane The crude nuclear fraction (4000 g/min pellet) separated from the liver homogenate by classical differential pelleting contains, in the case of liver, not only nuclei but sheets of plasma membrane, partially broken cells, aggregated material and large amounts of smaller organelles trapped in the pellet. These various structures are difficult to separate from each other by differential pelleting but may be separated successfully with an A-XI1 zonal rotor using a similar Siihieil s d r \ y . 287
210
200-
-.C 0 c 0 0
t
-2
100.
.-al C
c
2
a.
Fraction No. Fig. 7.2. Separation of particles from a rat liver homogenate in an A-XI1 zonal rotor. The sample comprised 20 ml of a homogenate (equivalent to 4 g of liver) prepared in 0.25 M sucrose, 5 mM NdHC03 pH 7.5 by 3 strokes of a PotterElvehjem homogeniser. The rotor was loaded with a complex sucrose gradient (Hinton 1972). The sample was injected and overlaid with 50 ml of 0.08 M sucrose. During acceleration, a further 70 ml of overlay solution was taken into the rotor. Centrifugation for 60 min at 3,700 revslmin at 4 (from Hinton 1972).
gradient to that used in the fractionation of the whole homogenate (Hinton et al. 1971; Hinton 1972. nb The molarity of the sucrose solution 'a' is wrongly printed as 1.164 in the first reference, the correct figure is 1.25 M. The density given for this solution and all other figures are correct. We would, however, now strongly recommend that as suggested in the second reference 0.3 M sucrose be used as the starting solution in place of the 0.25 M sucrose specified in the first reference). Apart from the nuclei themselves, the most interesting components of the crude nuclear fraction are the sheets of plasma membrane. If these are to be purified, most of the red blood cells must be removed before the crude nuclear fraction is prepared. This may be done either by perfusing the tissue prior to homogenisation or by homogenising
Ch. 7
APPLICATIONS
21 1
in a hypotonic medium. If a significant number of intact red blood cells are present in the 4000 g/min pellet, they will aggregate with plasma membrane sheets (Hinton 1972). Similar methods have been used for separating kidney plasma membranes (Price et al. 1972) and could, no doubt, be used for the separation of membranes from many other types of tissue where the intercellular links are strong enough to maintain the plasma membranes as large sheets. These methods cannot be used for the separation of plasma membranes from tissues which have weak intercellular links as in this case the cells are torn apart during homogenisation resulting in plasma membrane fragments which are too small to be separated from mitochondria by rate sedimentation (Hinton 1972). This is the case with many transplantable hepatomas (Prosper0 and Hinton 1973). The single step procedure discussed above depends on the use of an A-XI1 zonal rotor. If this is not available, a two-stage procedure must be used to separate plasma membrane sheets. Firstly microsomes are removed either by repeated washing or by sedimentation in a B-type zonal rotor. Simple washing is an effective procedure here as there is a very large difference in size between plasma membrane sheets and microsomes (see 9 3.1). Plasma membrane sheets are then separated from mitochondria, nuclei and other large components of the nuclear fraction by isopycnic flotation. These procedures were originated by Neville (1960) and Emmelot et al. (1964) and have since been applied to a large number of other tissues (Hinton 1972; de Pierre and Karnovsky 1973). 7.2.1.1. Purified nuclei As nuclei tend to aggregate on pelleting in 0.25 M sucrose, nuclei for use in subfractionation studies must be separated directly from the homogenate by pelleting through 2.2 M sucrose. Purified nuclei may then be fractionated on the basis of their size by rate sedimentation in an A-XI1 zonal rotor (Johnstone et al. 1968; Johnstone and Mathias 1972). The number of subfractions will depend on the tissue and the age and species of animal used. Fractions corresponding to diploid, tetraploid, octaploid and hexadecaploid nuclei can be separated from the livers of old mice (Fig. 7.3) .7lll>/'?I ,,vk.\ p 287
212
DENSITY GRADIENT C'ENTRIFUCiATION
To P
Effluent vol (mt)
Bottom
Fig.7.3. Separation of liver nuclei from mice of the NIH strain. Nuclei were purified by pelleting through 2.2 M sucrose. The purified nuclei were resuspended and layered over a 20-50% w/w sucrose gradient in an A-XI1 zonal rotor. The gradient contained 1 mM Mg Cll and was adjusted to pH 7.4 with NaHCO,. Centrifugation was for 1 h at 600 revs/min. The sharp peak at 200 ml shows the position of the sample. Further zones, from left to right, show diploid, tetraploid, octaploid and hexadecaploid nuclei (from Johnston and Mathias 1972).
and the results are an elegant illustration of the resolution which can be obtained with rate-zonal centrifugation. With less homogeneous tissues than liver, more complex patterns will be obtained as the nuclei will differ in size and shape (Austoker et al. 1972). Such methods may prove most valuable in exploring the mechanism by which different parts of the genome become activated during morphogenesis. 7.2.2. Subfractionation of the mitochondrial fraction The mitochondrial fraction separated from liver by differential pelleting contains, in addition to mitochondria, a variety of other subcellular components. Important among these are large sheets of endoplasmic reticulum such as surround mitochondria in many living cells and adhere to mitochondria after cell breakage unless chelating agents are included in the homogenisation medium (Chappel and Hansford 1972), lysosomes, peroxisomes and, if the tissue was homogenised gently, the Golgi apparatus. Some fragments of the plasma membrane may also be present, especially in homogenates of tissues
Ch. 7
213
APPLICATIONS
where intercellular bonds are weak. Secretion vacuoles may also be present (see Fleischer and Packer, 1974). Subfractionation of the mitochondrial components, or examination of the possible heterogeneity among mitochondria themselves or separation of fragments of broken mitochondria. Fig. 3.1 indicates that of the other cell components that are similar in size to mitochondria, and hence sediment in the mitochondrial fraction, lysosomes and peroxisomes are best separated by rate sedimentation, for their density is similar to that of rough endoplasmic reticulum fragments; fragments of the Golgi apparatus and of the plasma membrane, being heterogeneous in size, are best separated by isopycnic banding. Fragments of the Golgi apparatus band at a density of about 1.12 and may be readily separated from other components of the mitochondrial fraction (Morre et al. 1970, 1972, 1974). Plasma membrane fragments band at a density of 1.16 (Prospero and Hinton 1972) as compared with a density of 1.18 for undamaged mitochondria (Beaufay et al. 1964) and cannot be completely separated unless the density of the latter is increased by the addition of 2 mM CaCl, to the homogenisation medium (Prospero andHinton 1972).Thechoiceof rotor and gradient in such experiments will depend on the objectives (see § 3.5). If this is simply to isolate a single structure, then a step gradient in a swing-out rotor may be most convenient (Morre et al. 1972). Although mitochondria are remarkably uniform in density, they are heterogeneous in their sedimentation coefficients and can be subfractionated by rate sedimentation. Equally good results are achieved with the dynamically loaded A-XI1 zonal rotor (Swick et al. 1967), the reorienting SZ - 14 zonal rotor (Wilson and Casciato 1972) or with a swing-out rotor (Storrie and Attardi 1973; see Fig. 7.4). The choice depends on the amount of material to be fractionated. There have been reports that, contrary to what is stated above, mitochondria can be divided into two subfractions by isopycnic banding (Pollak and Nunn 1970); these results can, however, probably be explained by the greater susceptibility of some mitochondria to high hydrostatic pressure (Wattiaux 1974). Siih,ccr
iiidc,T
p 287
214
DENSITY GRADIENT CENTRIFUGATION
r
1
I
Fraction No. Fig. 7.4. Subfractionation of Hela mit;chondria. A crude mitochondria1 fraction, derived from a homopenate of 2-4 x 10 cells labelled for 20 min with S'H-uridine, was layered on a linear density gradient extending from @-200/, dextran and containing 0.48 M sucrose and 0.1 mM Tris pH 6.7 (25 ) in the tube of a SW-27 swing-out rotor (Beckman-Spinco). A cushion of 1.7 M sucrose, 0.01 M Tris pH 6.7 was placed at the bottom of the tube. Centrifugation was for 2 hr at 7,000 revs/min. Note the separation of mitochondria synthesising RNA and the mitochondria containing malate dehydrogenase (MDM) activity from those containing cytochrome oxidase (From Storrie and Attardi, 1973) N.B. In this figure the direction of sedimentation is from right to left.
Mitochondria are extremely complex structures, surrounded by two distinct membranes and possessing distinctive DNA and ribosomes. The mitochondrial inner and outer membranes have been found to have markedly different densities and to be readily separable by isopycnic banding on sucrose gradients (Sottocasa et al. 1967; Schaitman et al. 1967), while the inner membrane may itself be split into fragments of differing density (Werner and Neupert 1972). Methods for separating mitochondrial DNA and ribosomes are mentioned in Ch. 3 and cj 7.2.5. 7.2.3. Subfractionation of the lysosomal fraction Purification of lysosomes is extremely difficult (Beaufay 1969 ; Reid 1972) because of the inherent inefficiency of differential pelleting
Ch. 7
215
APPLICATIONS
(9 1.2.2) the 'classical' L or lysosomal fraction is neither pure nor, necessarily, representative. We feel that it is better to start with an M + L fraction prepared by centrifuging the post-nuclear supernatant for about 150,000 g/min. Lysosomes may be separated from such a fraction by rate zonal sedimentation (see Fig. 3.10); either a zonal o r a swing-out rotor may be used (Fig. 3.6). The lysosomal fraction prepared by rate sedimentation will be free from mitochondria but will contain microbodies, which are of the same size as lysosomes (see Fig. 3.1.), a few plasma membrane fragments and large densely-coated fragments of rough endoplasmic reticulum which tend to co-sediment with lysosomes (M. Dobrota and J.T.R. Fitzsimons, unpublished experiments). A partial separation of these components from the lysosomes may be achieved by isopycnic banding (Fig. 7.5). This is most conveniently carried out by loading the whole fraction separated by rate sedimentation onto a short isopycnic gradient (Dobrota and Hinton 1974). This procedure circumvents problems connected with the fragility of lysosomes on pelleting and resuspension. Lysosomes separated in this way are still contaminated with endoplasmic reticulum vesicles. These cannot be removed by centrifugal methods and if very pure lysosomes are required they must be separated by free-flow electrophoresis (Stahn et al. 1970; Henning et al. 1974). Lysosomes are an extremely heterogenous class of organelle. In addition to the distinction between primary lysosomes originating from the Golgi apparatus and secondary lysosomes, formed by fusion of primary lysosomes with phagosomes or autophagic vacuoles, some types of cell, notably circulating phagocytes, manufacture more than one type of primary lysosome. Secondary lysosomes may readily be separated from tissue homogenates after the injection of compounds suchas Triton WR-1339 (Wattiaux et al. 1963), dextran, iron-sorbitolcitrate complex (Arborgh et al. 1973) o r colloidal gold (Henning and Plattner 1974). These componds are taken up into lysosomes and modify their density so that they may be separated from other cell components by isopycnic banding (Beaufay 1969; Reid 1972). It is much more difficult to separate the different types of granules S,,h,'YI ,,,,I',\
p 287
216
DENSITY GRADIENT CENTRIFUGATION
FRACTION No
FRACTION No
Fig. 7.5.Isopycnic banding of lysosomes. A B-XIV zonal rotor was filled with 400 ml of sucrose gradient. The sample was 150 ml of the lysosome-rich region separated by centrifugation in an HS zonal rotor. Centrifugation for 150 min at 47,000 revs/ min. ---, density. Upper diagram, 0-0, acid phosphatase; A-A, acid acid phosphodiesterase; +p galactosidase. ribonuclease; v-v, Lower diagram, __ , protein; -0, 5’-nucleotidase; L O , glucose-6phosphatase; xx catalase; and, to facilitate comparison, 0-0, acid phosphatase.
+,
present in normal liver. No separation is achieved on isopycnic banding, but some separation can be obtained by rate sedimentation (Rahman et al. 1967; Burge and Hinton 1971). Canonico and Bird
Ch. 7
APPLICATIONS
217
(1970) achieved a similar fractionation of lysosomes from muscle; they interpret their results in terms of the separation of lysosomes from different cell types. As mentioned earlier, circulating phagocytes may possess more than one type of granule containing hydrolytic enzymes ; these may be separated by rate sedimentation (Baggiolini et al. 1969). Sometimes additional resolution may be achieved by isopycnic banding of the fractions separated by rate sedimentation (Baggiolini et al. 1970). Attempts have been made to separate the membranes of Triton WR- 1339-loaded secondary lysosomes (Thines-Sempoux 1973) and of ‘normal’ lysosomes from rat liver (Dobrota and Hinton 1974) by isopycnic banding after the membranes have been ruptured by vigorous homogenisation. However it is not clear whether the fractions so separated contain only the lysosomal membrane or whether they also contain the postulated lysosomal matrix. Similar comments apply to membranes prepared from lysosomes separated by free flow electrophoresis (Henning et al. 1974) although here the risk of contamination of the preparations by membranes deriving from other subcellular structures is much reduced.
7.2.4. Subfract ionat ion of microsomes The microsomal fraction is highly heterogeneous. The term rnicrosomes simply indicates the small particulate fraction which will not pellet at 150.000 g/min but is pelleted at about lo7 g/min. The subcellular source of these fragments depends on the tissue. In the case of liver, fragments of the rough and smooth endoplasmic reticulum are the major components, but fragments of the plasma membrane and of the Golgi apparatus are also present (Fleischer and Fleischer, 1970; Hinton 1972) as well as outer membrane fragments from broken mitochondria (Amar-Costesec 1974) and small lysosomes. As normally prepared, the microsomal fraction will also contain the free polysomes and glycogen particles together with a proportion of free ribosome subunits and of ferritin particles. The proportion but not the number of the various constituents will vary between different cell types. Sirhlrcr nidcr 11 287
218
DENSITY GRADIENT CENTRIFUGATION
Clearly the fractionation of such a complex mixture is not simple. The dense ribonucleoprotein particles, glycogen and ferritin granules may be separated by centrifugation through a layer of 2 M sucrose, although care must be taken to avoid the small particles tangling in the layer of membrane which forms over the 2 M sucrose. This may be avoided by using liver from fasted animals, by fractionating very dilute suspensions (Cow et al. 1970) or by recentrifuging the microsomes in 2 M sucrose in which case the membranes and the small particles will move in opposite directions. Glycogen granules can be fractionated according to their size by rate sedimentation on sucrose gradients and then separated from co-sedimenting contaminants by banding on CsCl gradients (Barber et al. 1966). However, it is not possible to fractionate the membrane-bound vesicles which make up the greater proportion of the microsomal fraction by rate sedimentation. A better separation can be achieved by isopycnic banding in sucrose gradients of a microsomal fraction prepared by differential centrifugation in either the tubes of swing-out rotors (see Tata 1972), highspeed zonal rotors (see Norris et al. 1974) or the special rotors developed fur isopycnic banding by Beaufay (Amar-Costesec et al. 1974). Similar results are achieved in all cases. Marker enzymes for the varioud membranous components of the microsomal fraction show very broad and overlapping distributions (Fig. 7.6).* One should be very careful about using non-linear gradients for this type of work as apparent separations may be achieved which are, in fact, gradient artefacts (Ch. 8). Thus in the experiment illustrated in Fig. 7.7, mathematical analysis shows that the apparent separation achieved on the step gradient is, in fact, exactly the same as the separation achieved on the linear gradient. Better separations between a proportion of the plasma membrane fragments and the endoplasmic
*
In the experiment illustrated in Fig. 7.6 the sample was applied at the dense end of the gradient and the particles separated by isopycnic flotation. A very similar pattern is obtained on isopycnic sedimentation although the degree of separation between the different types of particle is somewhat less.
Ch. 7
219
APPLICATIONS
1.2 I I I I
I
6 m
->
'2 0" 4.
1.1
20
10
30
1.0
Fraction no.
30 // / I / - ] 1.2 / /
'OI >
4-
>
P
2
15
4-
c
0 Fraction no.
Fig. 7.6. Subfractionation of microsomes by isopycnic flotation. A microsomal fraction. prepared from a rat liver homogenate by differential pelleting, was resuspended in I .8 M sucrose and loaded under a sucrose gradient in a B-XIV zonal rotor. Centrifugation was for 150 min at 47,000 revsimin. The gradient contained 5 mM Tris pH 7.4. (from Hinton et al. 1971). Sl,l,jl,'~lesk*\ S,,I~,?'l bsk.\. 1'. 287
220
DENSITY GRADIENT CENTRIFUGATION
Fig. 7.7. An illustration of the problems of interpreting separations on non-linear gradients. A microsomal fraction from rat liver was centrifuged for 3 hr at 40,000 revs/min in a B-IV zonal rotor in a gradient containing 5 mM MgC12 using (A) a linear gradient and (B) a gradient containing a flat step between two linear portions. and glucose-6-phosphatase (0-.-.-.0) In B zones rich in 5’-nucleotidase (0-0) appear to be clearly separated. However when the results from B are processed to show the density distribution of the fragments (C), the separation is nearly identical to that obtained in experiment A (from Prosper0 1973).
reticulum fragments is achieved when isopycnic flotation is used in place of isopycnic sedimentation, but the plasma membrane fragments which sediment with the microsomal fraction are themselves heterogeneous (Norris et al. 1974 ). With microsomes from Ehrlich ascites cells, much better results are achieved by isopycnic banding on Ficoll gradients than are achieved with sucrose gradients (Wallach 1967). No systematic comparison appears to have been carried out with liver microsomes,
Ch. 7
22 1
APPLICATIONS
but brief reports that plasma membrane and endoplasmic reticulum fragments form sharp bands after isopycnic banding on Ficoll gradients (House and Weidemann 1970) as against the rather broad (Hinton et al. 1971;Norris et al. 1974) bands obtained after separating in continuous sucrose gradients suggests that Ficoll may have advantages for the isopycnic banding of liver microsomes. Other high molecular weight solutes may be equally useful, as demonstrated by Graham (1972, 1973) who used Dextran gradients for the isopycnic separation of enzymically distinct membrane fragments from cultured kidney cells. While the different particles which make up the microsomal fraction can be partially separated by isopycnic banding in sucrose or Ficoll gradients, the fractions separated are still very heterogeneous. When the entire post-lysosomal supernatant is to be applied to a density gradient, the separation of endoplasmic reticulum fragments from other membraneous particles may be improved by the inclusion of Mg2+ions in the density gradient (El-Aaser et al. 1966). The degree of separation is not altered by changing the Mg concentration at least at concentrations greater than 1 mM (Hinton et al. 1967). The effect would appear to be due to attachment of free ribosomes to 'smooth' endoplasmic reticulum membranes, for the addition of Mg2+ ions to a suspension of microsomes has little effect on the density of the particles. The addition of Pb2+ions, in very low concentrations does, however, cause a large increase in density of endoplasmic reticulum fragments, and, providing that the aggregates which form are dispersed by sonication, does permit the separation of plasma membrane fragments from endoplasmic reticulum-derived vesicles (Hinton et al. 1971). This potentially powerful method for isolating one particular component of the microsomal fraction by using cytochemical methods to bind heavy metal ions to one type of membrane (Leskes et al. 1971) uses the resultant increase in density of that particular component to obtain its separation by isopycnic banding. Care must be taken to remove the non-specifically bound lead, if specific separations are to be obtained (Hinton et al. 1971). Enzymes do not seem to be +
Stihwr rrr&\
p 287
222
DENSITY GRADIENT CENTRIFUGATION
distributed evenly over the surface of microsomal fragments derived from the endoplasmic reticulum. Both rough and smooth microsomes can be broken into smaller fragments by sonication. These fragments may be separated by rate sedimentation on sucrose gradients in the tubes of swing-out rotors. Enzymes associated with the oxidation of NADH' show a different distribution to those connected with the oxidation of NADPH (Dallner et al. 1972). This suggests some mosaicism in the membrane.
7.2.5. Fractionation of' r ibonucleoprotein particles The fractionations of polysomes and of ribosome subunits by rate sedimentation on sucrose gradients were among the earliest applications of the technique (McQuillen et al. 1959). Separations may be carried out either on swing-out (No11 1969) or zonal rotors (Birnie 1972). The separation of polysomes, in particular, is one of the best tests for technique in density gradient centrifugation, as polysomes form a nicely graded series in which each particle is slightly more similar in size to the next largest than it is to the next smallest. Thus, trimers (154 S) are 25% larger than dimers (123 S) but tetramers (183 S) are only 18% larger than trimers and so on (Puderer et al. 1965). Polysomes up to the 12-mer have been resolved as separate peaks (Norman, 1971) but most workers are satisfied if they can separate up to the 7-mer or 8-mer. Density gradient centrifugation can also be used to reveal slight differences between newly formed small ribosome subunits and ribosome subunits recycled from polysomes and to fractionate ribonucleoprotein particles extracted from the nucleoplasm (Lukanidin et al. 1972) or the nucleolus (Prestayko et al. 1972). Ribosome subunits may also be fractionated according to their density by isopycnic banding in CsCl gradients (Perry and Kelley 1966), provided that they have previously been fixed with formaldehyde. This technique is useful for exploring heterogeneity among the subunits and in separating mRNA-containing particles from dissociated polysomes (Henshaw 1968) (see Fig. 7.8) or from the other small ribonucleoprotein particles sedimenting free in the cytoplasm
Ch. 7
223
APPLICATIONS
300
Tdbe number
Fig. 7.8. CsCl equilibrium gradient centrifugation of rat liver ribosome subunits. Rat liver polyribosomes were treated with EDTA and fixed with 6:; formaldehyde. Separation was by centrifugation for 64 hr at 40,000 revs !I an a Spinco SW 50 swing-out rotor. The peak at density 1.46 represents mRNA-containing particles. The peaks at about I .52 and I .59 are due respectively to small and large ribosomal x , AZbOnn1; 0-----C. RNA labelled for 40 min in vivo with subunits. x ''C-orotic acid: 00 . RNA labelled for 44 hr in vivo with 'H-orotic acid (from Henshaw 1968). ~
(Henshaw et al. 1967; Ayuso-Parila et al. 1973). One must, however, take great care in interpreting results obtained from ribonucleoprotein particle preparations contaminated by any significant amount of cytoplasmic protein (Fig. 7.9). A considerable amount of protein may be bound during fixation resulting in a much lower banding density than would have been obtained with the pure particles. Furthermore, recent experiments (McConkey 1974) have cast some doubt on whether there is the strict relationship, hitherto assumed, between the density of ribonucleoprotein particles in CsCl and their chemical composition. Unfixed ribonucleoprotein particles may be banded on Metrizamide SuhpI
lltde\
p 287
224
DENSITY GRADIENT CENTRIFUGATION
A200 nm
0.Sr
Fraction no. A200 nm
I
Fraction no.
Fig. 7.9. Artefacts introduced by the presence of high concentrations of extraneous proteins during fixation. Polysomes were separated by the method of Leitin and Lerman (1970) and fixed with 6% formaldehyde buffered according to the method of Perry and Kelley (1968) either a) alone or b) in the presence of about 1 mg/ml of rat liver cytosol protein. Centrifugation was for 18 hr at 50,000 revs/min in an MSE 3 x 6.5 ml swing-out rotor using preformed CsCl gradients. (B.M.Mullock and R.H. Hinton, unpublished experiment).
gradients (Mullock and Hinton 1973; Hinton et al. 1974b), and Buckingham and Gros succeeded in separating mRNA-containing particles. Ribonucleoprotein particles have lower densities in Metriz-
Ch. 7
APPLICATIONS
225
amide solutions (Hinton et al. 1974) than in either sucrose (Petermann 1964) or CsCl gradients (Perry and Kelley 1965). This could be due to variation in the degree of hydration of the RNA and protein molecules (Hinton et al. 1974). A similar dependence of apparent density with medium is observed with chromatin (see Chs. 2 8z 3). 7.2.6. Fractionation of chromatin Probably the greatest gap in our knowledge of cells is our lack of understanding of the processes which control the development of the tissues of higher organisms. These changes may be largely expressed by means of synthesis of new messenger RNA to code for the novel proteins needed for the development of specialised tissue. Comparison of the structure of active and inactive regions of chromatin should shed some light on the processes by which the many segments of DNA needed to code for the tissue specific proteins may be activated. There has, therefore, been a rapid increase in interest in methods for fractionating intact Aromatin. Most centrifugal methods for preparing or fractionating chromatin depend on isopycnic banding, although Bhorjee and Pederson (1973) have used rate edimentation through concentrated sucrose gradients for separating chromatin from the smaller ribonucleoprotein particles. Chromatin is too dense to band in sucrose gradients and is dissociated by CsCl (MacGillivray et al. 1972). Unfixed chromatin may be banded in sucrose-glucose (Raynaud and Ohlenbusch 1972) or in chloral hydrate gradients (Hossainy et al. 1973). However, the most suitable medium for banding chromatin appears to be Metrizamide, in which separation of chromatin from other nuclear constituents (Birnie et al. 1973a) and subfractionation of sheared chromatin (Rickwood et al. 1974) have been demonstrated. Chromatin isolated by such methods contains RNA in addition to DNA and protein. A simple method for separating these three classes of macromolecule has been devised by Monahan and Hull (1974). The chromatin is treated with the detergent sarcosyl (sodium dodecyl sarcosinate) and centrifuged to equilibrium on a gradient of Cs,SO, containing sarcosyl and dimethylsulphoxide. Siihiecl a d c r p 287
226
DENSITY GRADIENT CENTRIFUGATION
7.3. Separation of subcellular structures from plant cells The presence of a tough cell wall makes difficult the breakage of plant cells without extensive damage to intracellular structures. This has meant that much less work has been carried out on the separation of plant subcellular structures than is the case with mammalian tissues. In the following paragraphs we shall outline only the separation of the larger subcellular structures. Methods for the separation of polysomes, ribosomes and ribosome subunits from plant cells resemble those used with mammalian cells (e.g. Beevers and Poulson 1972; Jones et al. 1973) but allow for differences in the homogenisation medium and require considerable precautions against RNA degradation. When separating plant ribosomes it is necessary to remember that, as well as the cytoplasmic ribosomes, the mitochondria and chloroplasts have their own protein synthesis systems which include distinctive ribosomes. Electron microscopy of plant tissue is more difficult than of mammalian cells and establishment of markers for the various subcellular structures has proved much more difficult. Some of the problems are discussed by Halliwell (1974). Further confusion has been introduced by the giving of names to imperfectly characterised structures. Thus glyoxysomes and microbodies have been spoken of as two distinct structures when, in fact, the former appears to be only a specialised form of the latter (de Duve 1969). When one eliminates ‘doubles’ of this type it becomes evident that plant cells contain the same range of structures as mammalian cells, namely nuclei, mitochondria, lysosomes (although in a very adapted form), microbodies, Golgi apparatus, endoplasmic reticulum and plasma membrane. The only novel structures are the chloroplasts and their precursors. Some tissues also contain storage granules (Schnarrenberger et al. 1972b) and zymogen granules (Cohen et al. 1971). The largest subcellular structures, apart from nuclei, in plant cells are the chloroplasts. The outer membrane of the chloroplasts of spinach leaves (Rocha and Ting 1970) and of other tissues is easily broken during homogenisation or pelleting. Intact chloroplasts can,
Ch. 7
APPLICATIONS
221
however, be separated by rate sedimentation in an A-XII* zonal rotor (Still and Price 1967) or in the tubes of a swing-out rotor (Rocha and Ting 1970). Intact spinach leaf chloroplasts have a sedimentation coefficient of 200,000 S or more and band in sucrose gradients at a density of 1.21. Broken chloroplasts sediment more slowly and band at a density of 1 .I7 (Rocha and Ting 1970). While the mature chloroplasts may be easily recognised from their chlorophyll content, it is more difficult to recognise their proplastid precursors. Proplastids from the cotyledons of sunflower seeds germinated in the dark band at a density of 1.26 along with microbodies, but retain a characteristic enzyme content (Schnarrenberger et al. 1972a). Plant mitochondria and microbodies are most readily separated from broken chloroplasts by isopycnic banding on sucrose gradients. Excellent results are obtained both with spinach leaf homogenates (Rocha and Ting 1970; Donaldson et al. 1972) and with sunflower cotyledons (Donaldson et al. 1972). An example of such a separation is shown in Fig. 7.10. It should be noted that the separation, as shown, depends on the chloroplasts being broken either by homogenisation or during pelleting. Intact chloroplasts are of the same density as mitochondria, but may be removed by rate sedimentation. Plant mitochondria band in sucrose gradients at a density of 1.21, microbodies band at a density of 1.25 (Rocha and Ting 1970; Donaldson et al. 1972). Spinach leaf mitochondria have a sedimentation coefficientoftheorderof 3000 S, the microbodies a sedimentation coefficient of the order of 9000 S (Rocha and Ting 1970). The evidence for the existence of lysosomes in plants has been reviewed by Matile (1969) and Gahan ( I 973). Both authors conclude that, with plant cells one must think in terms of a system, corresponding to the Golgi-endoplasmic reticulum-lysosome system postulated in animal cells (Novikoff et a]. 1971). In plants, what are thought of as typical lysosomal enzymes are found in the endoplasmic reticum, in the dictyosomes of the Golgi apparatus, in pre-vacuolar vesicles * The experiments described in the reference were actually carried out using an A-IX zonal rotor, an earlier form of the A-XI1 zonal rotor. S u h p I w d e ~11 287
228
DENSITY GRADIENT CENTRIFUGATION
41r=4 sucre..
g a'
20
10
I
n
Volume Lmll
Fig. 7.10. Distribution of subcellular organelles from spinach leaves. A large particulate fraction was prepared by differential centrifugation and then further fractionated by centrifugation for 2 hr at 30,000 revs/min in a B-XXX zonal rotor. Markers are as follows: chlorophyll for chloroplasts, catalase for microbodies and cytochrome oxidase for mitochondria. NADH-cytochrome c reductase is present in both mitochondria and endoplasmic reticulum fragments (from Donaldson et al. 1972).
and in vacuoles. The last two structures would correspond to the lysosomes of mammalian cells, but play a much more elaborate part in the metabolism of plant cells than is performed by mammalian lysosomes. The small vactoles present in the meristematic root tip
Ch. 7
APPLICATIONS
229
cells of corn seedlings have been isolated by the isopycnic banding of a ‘mitochondrial’ fraction on a sucrose gradient (Matile 1968). The size, and consequent fragility, of the vacuoles of parenchymatous cells has so far prevented their isolation. Granules containing storage proteins have been isolated from sunflower cotyledons (Scharrenburger et al. 1972b) and from the reserve tissues of seeds (see Matile 1969). The latter granules appear to contain hydrolytic enzymes. ‘Zymogen bodies’ containing a starch debranching enzyme in inactive form may be isolated from pea seeds by isopycnic banding on a Ficoll gradient (Cohen et al. 1971). Plant microsomes have been little studied. The microsomal fraction as separated by differential pelleting is very heterogeneous, containing fragments of the endoplasmic reticulum, the Golgi apparatus and the plasma and vacuole membranes. In the absence of markers for structures other than the endoplasmic reticulum (Halliwell 1974) it is not possible to draw any definite conclusions about the purity of the various subfractions which may be separated by isopycnic banding (Lord et al. 1973).
7.4. Separation of subcellular components from unicellular organisms In bacterial cells the major role of density gradient centrifugation has been in the separation of the DNA or the ribosomes. DNA separation is discussed later (0 7.5.1.2). The techniques for separating bacterial polysomes, ribosomesandribosomalsubunitsarevery similar to the methods used with the corresponding structures from mammalian cells. It should be noted that very rapid cooling of the growing cells and careful lysis by digestion with lysozyme in the presence of a detergent is essential if intact polysomes are to be separated (Godson and Sinsheimer 1967; No11 1969). The separation of ribosomal subunits from bacteria is very simple. Up to 2 g of subunits can be separated at any one time by the use ofa steep sucrose gradient (Eikenberry et al. 1970; Sypherd and Wireman 1974). Free-living eucaryote cells have at least as wide a range of intraSuhlerf index p 287
230
DENSITY GRADIENT CENTRIFUGATION
cellular structures as mammalian and plant cells. Like the latter, the cells of free-living eucaryotes are surrounded by a strong cell wall and one of the major difficulties in separating the intracellular structures is to break the cell wall without simultaneously breaking all the intracellular membranes. Another similarity with plant cells is the difficulty of finding reliable markers for the various intracellular structures. Recently, however, a comprehensive discussion of possible marker enzymes for the larger subcellular structures of a wide range of eucaryotic micro-organisms by Lloyd and Cartledge (1974) has partly clarified the field. The most commonly used method for fractionating homogenates from protozea and yeasts has been isopycnic banding on sucrose gradients. In most cases, work has been concentrated on the larger subcellular particles. The following short summary is taken from the article by Lloyd and Cartledge (1974) which should be consulted for further references. Among protozoa, the most extensive work on the isolation of subcellular structures has been with the ciliate Tetrahymena pyriformis. Lloyd et al. (1971) and Muller (1972) have reported the patterns obtained by isopycnic banding of the large particulate fraction on sucrose gradients. Mitochondria, microbodies and endoplasmic reticulum fragments were well separated, but lysosomal enzymes showed a very complex pattern which indicated the presence of at least three populations with distinctive enzymology. There is considerable cross contamination between the various fractions. Little work appears to have been carried out on the fractionation of Tetrahymena homogenates by rate sedimentation, but there are indications that it is possible to obtain a considerable degree of purification of lysosomes in this way (Lloyd et al. 1971; Cooper and Dobrota 1974). Other protozoa have been studied in less detail. The soil amoeba Hartmanella castellanii gave results very similar to those obtained with Tetrahymena pyriformis except that there was less separation between mitochondria and microbodies. A serious problem in all work on micro-organism is illustrated by the experiments of Wiener and Ashworth (1970) on the myxamoeba of Dictyostelium discoideum. The
Ch. 7
APPLICATIONS
23 1
isopycnic banding densities of the lysosomes and microbodies and their enzyme composition varied markedly with the conditions of growth. The parasitic flagellate Trichomonas foetus has been fractionated by isopycnic banding in the E 40 isopycnic banding rotor developed by Beaufay (Muller 1973). The results are interesting because of the unsuspected enzyme composition of what appears, morphologically, to be a fraction highly enriched in microbodies. These results emphasise again the necessity for establishing markers for each species of organism which is studied. The results obtained with yeasts are, in general similar to those obtained with protozoa. It has been possible to separate intact cells of the fission yeast Schizosaccharomyces pombe in different stages of growth by centrifugation in a HS zonal rotor using rate sedimentation on a sucrose or Ficoll gradient and by isopycnic banding in a B-XIV zonal rotor (Poole and Lloyd 1973). The centrifugal behavior of subcellular particles from Saccharomyces cerevisiae has been extensively studied by Cartledge and Lloyd (1972). On isopycnic banding, mitochondria were split into two subfractions, but it is not clear whether this is due to intrinsic heterogeneity or to damage such as can occur during the isopycnic banding of liver mitochondria in a swing-out rotor (Beaufay 1966). Yeast microbodies are poorly separated from mitochondria on isopycnic banding, but may be separated by rate sedimentation (Cartledge et al. 1971). Yeast ‘lysosomes’ are very heterogeneous in density and sedimentation coefficient. Clearly there are several different types of structure containing acid hydrolases, but it has not yet been possible to identify these with intracellular structures visible in the electron microscope. Subcellular fractionation of algae give results similar to those obtained with higher plants (Lloyd and Cartledge 1974). Little work has been carried out on the systematic subcellular fractionation of fungal cells although Neurospora crassa has been a favoured species for the study of mitochondria1 DNA. References to fractionation methods will be found in Lloyd and Cartledge (1974). Up till now, studies on eucaryotic microorganisms have concentrated on establishing the distribution of enzymes between the larger Stihli < I wI" (w/v) Na2S04, and redissolving in 10 mlO.9% NaCI. Swell 1 g cyanogen bromide (CNBr)-activated Sepharose 4B (Pharmacia AB) in 10 mM HCI. This should be carried out in a Buchner funnel with a grade 4 Scinta glass filter. A total of 200 ml of 10 mM HCI is added and removed in several aliquots. After a final wash with 0.1 M NaHCO, transfer the gel to a glass tube, add 50 ml of 0.1 M NaHCO, containing 1 ml of the y-globulin solution from (I), and mix by vertical rotation (25 revlrnin) for 24 hr at room temperature. Wash the immunosorbent once with 100 ml of 1.0 M ethanolamine. pH 8.0, with continuous mixing for 2 hr and then twice with 100 ml of 0.5 M NaHCO,, twice with 50 ml of 0.2 M sodium acetate buffer, pH 4.0, and twice with 100 ml of assay diluent (identical with that shown in Table 1.2, but containing in addition 0. I :{, v/v Tween-20 (Koch-Light Laboratories Ltd.), with 30 min continuous mixing for each wash. The standard curve is set up as shown in Table 1.3, but substituting a suspension of the solid phase for the soluble antibody. The appropriate dilution of the solid phase will vary between different antisera and must be established by experiment with different dilutions. Cap all tubes and mix by vertical rotation (approx. 25 revimin) for I hr at room temperature. Centrifuge at 2000 x g for 5 min, uncap the tubes, aspirate the supernatants leaving a constant volume (0.1 ml) in each tube, and wash once with 1 ml assay diluent. Estimate the counts in the solid phase by placing each tube in the wellcrystal of a y-counter. Calculate results (Table 1.2).
Ch. 5
SEPARATION OF B O U N D A N D FREE LlCiAND
421
applied to virtually any binder capable of covalent attachment to the particle; (2) they are highly efficient and produce virtually complete separation of the bound fraction; (3) they give excellent precision if carefully used; (4) they are not as liable as some other systems to non-specific effects introduced by plasma and serum. At the same time, solid-phase methods have certain disadvantages which explain why these sophisticated systems are not in universal use: (1) the primary reagent is tedious to prepare; (2) the recovery of antibody activity on the solid phase is only 30% or less of that in the original y-globulin preparation, probably due to the fact that many of the molecules attach to the particles through their combining sites; the waste of activity is only acceptable if the supply of antiserum is abundant; ( 3 ) in the case of antibodies to larger molecules, attachment to solid phase results in a loss of affinity and hence of sensitivity in the assay; this is only critical with that minority of assays in which extreme sensitivity is required; (4)finally, and most important, the actual assay procedure is more complex than with some other systems; for continuous mixing the tubes must be capped and uncapped; a t the end of the procedure the washing of the solid phase sometimes involves several steps of centrifugation and aspiration. Thus, for any assay with a large sample throughput, particle solid-phase systems are not technically convenient. 5.3.7. Conclusions : the choice of'u separation procedure There is no magic in any one separation procedure. For a given assay system the choice potentially embraces all the techniques described above, and there is no reason to follow slavishly an earlier published technique. Instances abound of the perpetuation of a complex and sometimes inefficient method simply because it was the first to be described. For example, charcoal separation continues to be used in assays for steroids and small peptides, systems in which dissociation of the bound complex is very likely to occur and thus impair efficiency. Other systems offering equal and probably greater convenience are available and should be used. There is much to be said for keeping the number of separation Suh,r 0.1 mCi) should be carried out in a tray which is monitored immediately after use. Where possible all glassware should be disposable. (10) All non-disposable equipment (columns, syringes, pipettes) used for high levels of activity should be thoroughly rinsed in running water after use and then allowed to soak in a detergent solution until required again. Such equipment should always be stored in the 'hot' laboratory and not taken for use outside. ( I 1) All tools (bottle openers. forceps) should be checked for contamination before being put away. (12) All sources of activity greater than 0. I mCi should be stored in a separate and designated area. preferably with lead shielding. Moriiioring o/ iIw luhoraiorjs All benches and equipment in low activity laboratories should be monitored once a week. and i n high activity laboratories after completion of every procedure. The permissible level of contamination is 10-'pCi/cmz averaged over an area not greater than 300 cmz. Monitoring o/ pi~rsoritrrl All personnel should have a general medical examination and a full blood exdmination at the time of first employment. The blood examination should be repeated at yearly intervals. All working staff should be provided with a film-badge which is examined and changed at monthly intervals. The maximum permissible dose level Suhjwr d i * rp. 5.11
526
RADIOIMMUNOASSAY A N D RELATED TECHNIQUES
is equivalent to 5 rem* per year to critical organs such as the gonads, bone marrow, and eye, or 3 rem in any one-quarter. In practice, such doses are very rare even in the busiest radioassay laboratory. Personnel working with iodine isotopes should have a thyroid count performed every month. Disposal of’ radioactive waste The rules for disposal of radioactive waste vary widely both between countries and within the same country. For this reason every laboratory should consult their appropriate local authority for guidance. Emergency procedure .for spills of’ radioactive materials The RSO and Head of Department should be informed at once. No personnel should be permitted into the affected area unless wearing gloves and overshoes, both of which must be disposed of on leaving the area. Any contaminated clothing must be removed and discarded. If the spill is on the skin the area should be flushed thoroughly with tap water taking great care not to spread the contamination. Decontamination should aim at getting activity down to pCi/cmz. Contaminated areas should be carefully mopped with disposable paper, then washed until activity is down to pCi/cm*. If this cannot be achieved the area must be temporarily covered with polythene sheeting, until dealt with by the RSO. Desigri of’the radioassay laboratory Radioassay kits, as opposed to bulk isotopes, can be handled in any general clinical laboratory (class 3). However, their use should be confined to a specified area which is provided with appropriate warning signs and is regularly monitored for contamination. Bulk isotopes ( 1 mCi or greater) should only be handled in a class 2 radioisotope laboratory which has the following general specifications: (1) walls and ceilings: washable, non-porous paint; ( 2 ) floor: washable linoleum, rubber, or vinyl; joins between floors and walls should be rounded; (3) sinks: connected to main drainage; stoppers and taps should be operated by foot-pedal; (4) ventilation: air-flow must be to the outside of the building or to an appropriate filter; ( 5 ) fume cupboards: must have an eddy-free airflow drawing at least 150 ft/min and vented to the outside of the building or to an appropriate filter. A self-contained charcoal-filter unit with 90% trapping efficiency is available from Interex Corp., 3 Strathmore Road, Natick, Mass. 01760, U.S.A.; ( 6 ) shower: a shower facility should be available immediately adjacent to the high-activity area. The amount of isotope (if any) which can be disposed of via mains drainage is determined by local regulations.
* The ‘rem’ is a dose equivalent derived from the absorbed dose measured in ‘rad’, and a quality factor and a distribution factor which depend on the type of radiation. One rad is the dose absorbed when 62.5 x lo6 MeV of energy is deposited in 1 g of matter.
References ADDISON. G.M. and C.N. HALES (1971) Horm. Res. 3, 59. ANDERSON, D.C. (1970) Clin. Chim. Acta 29, 513. ANDRIE.~. J., S. MANASand F. DRAY(1975) In: E. Cameron, S. Hillier and K. Griffiths (eds.) Steroid lmmunoassay (Alpha-Omega-Alpha, Cardiff) p. 189. BANCHAM, D.R. and P.M. COTES(1974) Brit. Med. Bull. 30, 12. BARISAS, B.G.. S.J. SINGER and J.M. STURTEVANT (1977) Immunochemistry 14. 247. BERSON, S.A. and R.S. YALOW (1966) Science 152, 205. BERSON, S.A., R.S. YALOW, A. BALJMAN. M.A. ROTHSCHILD and K. NEWERLY (1956) J. Clin. Invest. 35, 170. BINOUX,M.A. and S.E. ODELL (1973) J. Clin. Endocrinol. Metab. 36, 303. BIRNBAUMER. L. and S.L. POHL(1973) J. Biol. Chem. 248,2056. BLOOM, S. (1974) Brit. Med. Bull. 30, 62. BOLTON, A.E. and W.M. HUNTER (1973) Biochem. J. 133, 529. BOYD,G.W., J. LANDON and W.S. PEART (1967) Lancet ii, 1002. BROOKER,G.. W.L. TERASAKI and M.G. PRICE(1976) Science 194,270. BUTT,W.R. (1972) J. Endocrinol. 55, 453. CATT,K.J. and C.W. TREGEAR (1967) Science 158, 1570. CATT,K.J., M.L. DUFALJ and T. TSLJRUHARA (1972) J. Clin. Endocrinol. Metab. 34, 123. CHARD,T. (1971) In: K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 491. CHARD, T., M.J. KITAU and J. LANDON (1970) J. Endocrinol. 46, 269. CHARD, T., M.J. MARTIN and J. LANDON (1971) In: K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 257. CHOPRA, I.J. (1972) J. Clin. Endocrinol. 34, 938. COLLINS. W.P. and J.F. HENNAM (1976) Mol. Asp. Med. I, 1 . DEAN,P.D.G., D. EXLEY and M.W. JOHNSON (1971) Steroids 18, 543. DESBUQLJOIS, B. and G.D. AURBACH (1971) J. Clin. Endocrinol. Metab. 33, 732. DONINI, S . and P. DONINI (1969) Acta Endocrinol. (Copenhagen) Suppl. 142,257. EKINS, R.P. (1960) Clin. Chim. Acta 5,453. EKINS,R.P. (1969) In: M. Margoulies (ed.) Protein and Polypeptide Hormones (Excerpta Medica, Amsterdam) p. 633. EKINS, R.P. (1974) Brit. Med. Bull. 30, 3. ERLANGER, B.F. (1973) Pharmacol. Rev. 25, 271. FISCHER, L. (1971) An Introduction to Gel Chromatography (North-Holland, Amsterdam). FRANCHIMONT, P. (1971) In: K. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 535. 527
Subjerr kdcr p . 531
528
RADIOIMMUNOASSAY A N D RELATED TECHNIQUES
FREEDLENDER, A. and R.E. CATHOU (1971) In: K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 94. FRENKEL; E.P.. S. KELLER and M.S. MCCALL (1966) J. Lab. Clin. Med. 68, 510. GALEN, R.S. and D. FORMAN (1977) Clin. Chem. 23, 119. GARDNER, J., G. BAILEY and T. CHARD (1974) Biochem. J. 137.469. GIBSON, R.G., B.I. HIRSCHOWITZ and A.A. MIHAS (1977) Clin. Chem. 23, 1046. GIESE,J. and M.D. NIELSEN (1971) In; K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 341. GILMAN A.G. . (1970) Proc. Natl. Acad. Sci. U.S.A. 67, 305. GORDON.Y.B., M.J. MARTIN, A.T. MCNEILE and T. C H A R D (1973) Lancet ii. 1168. GRANT, D.B. (1968) Acta Endocrinol. (Copenhagen) 5Y, 139. GREEN, I., W.E. PAULand B. BENACERRAF (1969) Proc. Natl. Acad. Sci. U.S.A. 64, 1095. GREENWOOD, F.C., W.M. HUNTER and J.S. GLOVER (1963) Biochem. J. 89, 114. GRODSKY. G.M. and P.H. FORSHAM (1960) J. Clin. Invest. 3Y, 1070. HABER, E., L.B. PACEand F.F. RICHARDS (1965) Anal. Biochem. f2, 163. HALES.C.N. and P.J. RANDLE (1963) Biochem. J. 88, 137. HALES,C.N.. P. BECK,M.J. EVANS and J.S. WOODHEAD (1975) In: C.A. Pasternak (ed.) Radioimmunoassay in Clinical Biochemistry (Heydcn, London) p. 283. HEDING, L.G. (1966) In: L. Donato, G . Milhaud and J. Sirchis (eds.) Labelled Proteins in Tracer Studies (EAEC, Brussels) p. 345. HERBERT, V., K.S. LAU,C.W. GOTTLIEB and S.J. BLEICHER (1965) J. Clin. Endocrinol. Metab. 25, 1375. HOLLANDER, F.C. den and A.H.W.M. SCHUURS (1971) In; K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 419. HUNTER, W. (1971) In: K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 3. HUNTER, W.M. and F.C. GREENWOOD (1964) Biochem. J. YI, 43. H U R N B.A.L. , and J. LANDON (1971) 61: K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone. Edinburgh) p. 121. HWANC,P., H. GUYDAand H. FRIESEN (1971) Proc. Natl. Acad. Sci. U.S.A. 68, 1902. JACOBY, B. and K.D. BAGSHAWE (1972) Cancer Res. 32.2413. JAMES,M.A.R.. T. CHARD and M.L. FORSLING (1971) In: K.E. Kirkham and W.M. Hunter (eds.) Radioimmunoassay Methods (Churchill-Livingstone, Edinburgh) p. 545. JOHANSSON, E.D.B. (1969) Acta Endocrinol. (Copenhagen) 61, 607. JOHNSON, E.G., T.E. STONEYCYPHER and B.E. STURGIS (1976) Clin. Chem. 22, 1164. KAGEN, L., S. SCHEIDT and L. ROBERTS (1975) Amer. J. Med. 58. 177. M. SIRENand U. SEUDERLING (1975) Anal. Biochem. KARONEN, S.-L., P. MORSKY, 67. 1.
REF t R E NC ES
529
KEANE, P.M.. W.H.C. W A L K E R J.,GAULD1EandG.E. ABRAHAM (1976) Chn. Chem. 22, 70. K O R E N M AS.G. N . (1968) J. Clin. Endocrinol. Metab. 28. 127. LANDESMAN, R. and B. SAXENA (1976) Fertil. Steril. 27, 357. LANDON, J., J. C A R N E and Y D. LANGLEY (1977) Ann. Clin. Biochem. 14.90. G. (1972) In: G.N. Vyas, H.A. Perkins and R. Schinid (eds.) Hepatitis LE BOUVIER. and Blood Transfusion (Grune and Stratton. New York) p. 97. LEFKOWITZ. R.J.. J. ROTHand J. PASTAN (1970) Science 170. 622. LEUTE,R., E.F. ULLMANand A. GOLDSTEIN (1972) J. Amer. Med. Assoc. 221. 1231. LEVINE, L. and R.M. GUTIERREZ-CERNOSEK (1972) Prostaglandins 2,281. MARCHAL.ONIS. J.J. (1969) Biochem. J. 113, 299. MARTIN.M.J. and J. LANDON(1975) In: C.A. Pasternak (ed.) Radioimmunoassay in Clinical Biochemistry (Heyden, London) p. 269. MCFARLANE, A S . (1958) Nature (London) 182. 53. MCGUIRE, W.L. (1973) J. Clin. Invest. 52. 73. MEDHI.S.Q. (1975) In: C.A. Pasternak (ed.) Radioimmunoassay in Clinical Biochemistry (Heyden. London) p. 21 3. MILES,L.E.M. and C.N. HALES(1968) Nature (London) ZIY, 186. MITCHELL, M.L., A.B. H A R D E N and M.E. O'ROURKE(1960) J. Clin. Endocrinol. Metab. 20. 1474. MITCHELL, M.L. and J. BYRON(1967) Diabetes 16, 656. MORGAN,C.R. and W.A. LAZAROW (1963) Diabetes 12, 115. MORGAN. C.R., R.L. SORENSON and A. LAZAROW (1974) Diabetes 13, 1 and 579. MURPHY, B.E.P., W. ENGELBERG and C.J. PATTEE(1963) J. Clin. Endocrinol. Metab. 23, 293. NIESCHLAG, E. (1975) Immunization with Hormones in Reproduction Research (North-Holland, Amsterdam). NYE. L., G.C. FORREST, H. GREENWOOD. J.S. GARDNER. R. JAY,J.R. ROBERTS and J. LANDON (1976) Clin. Chim. Acta 6Y, 387. ORSKOV, H. (1967) Scand. J. Clin. Lab. Invest. 20, 297. PORT-ER, R.R. (1959) Biochem. J. 73. 119. PREEDY, J.R.K. and E.H. A I T K E (1961) N J . Biol. Chem. 236, 1297. RATCLIFFE, J.G. (1974) Brit. Med. Bull. 30, 32. REDSHAW. M.R. and S.S. LYNCH(1974) J. Endocrinol. 60. 527. REES,L.H. (1976) Clin. Endocrinol. (Oxford) 5,3635. REES.L.H..D.M. COOK. J.W. KENDAI-L. C.F. ALLEN, R.M. KRAMER.J.G. RATCLIFFE and R.A. KNIGHT (1971) Endocrinology 8Y, 254. REES,L.H.,J.G. RATCLIFFE,G.M. BESSER. R.M. K R A M E R J. , LAND ON^^^ J. C H A Y E N (1973) Nature New Biol. 241, 84. RODBARD,D. (1971) In: W.D. Odell and W.H. Daughaday (eds.) Principles of Competitive Protein Binding Assays (Lippincott, Philadelphia. Pa.) p. 204. Sshjwi i1vle.r p . 531
530
RADIOIMMUNOASSAY AND RELATED TECHNIQUES
RODBARD, D. and J.E. LEWALD (1970) Acta Endocrinol. (Copenhagen) Suppl. 147, 79. F. PENNISI, N. RICCIONI, P. GIANONI and R. GIORDANI ROSA,U., G.A. SCASSELLATI, (1964) Biochim. Biophys. Acta 86, 519. ROSNER, W. (1972) J. Clin. Endocrinol. Metab. 34, 983. G., R. ASSAN,R.S. YALOWand S.A. BERSON(1969) Nature (London) ROSSELIN, 212, 355. RUBENSTEIN, K., R. SCHNEIDER and E. ULLMAN (1972) Biochem. Biophys. Res. Commun. 47,846. SAMOLS, E.C. and D. BILKUS(1963) Proc. SOC.Exp. Biol. Med. 115, 89. SCHARPE, E.L., W.M. COOREMAN, W.J. BLOMME and G.M. LAEKEMAN (1976) Clin. Chem. 22, 733. SELA,M. (1969) Science 166, 1365. A.H. RUBENSTEIN, S. CHO and C. BAYLISS(1968) STEINER, D.F., 0.HALLUND, Diabetes 17, 725. STAHL,E. (ed.) (1969) Thin Layer Chromatography (English edition) (Springer Verlag, New York). SUN,L. and V. SPIEHLER (1976) Clin. Chem. 22, 2029. K. and J. FERRIN (1968) J. Clin. Endocrinol. Metab. 28, 1667. THOMAS, THORELL, J.J. and B.G. JOHANSSON (1971) Biochim. Biophys. Acta 251, 363. TRAFFORD, D.J.H., P.R. WARD,A.Y. Foo and H.L.J. MAKIN(1974) J. Endocrinol. 65.44.
TYLER, J.P.P.. J.F. HENNAM, J.R. NEWTON and W.P. COLLINS (1973) Steroids 22, 871. ULLMAN, E., J. BLAKEMORE and R. LEUTE (1975) Clin. Chem. 21, 1011. UPDIKE,S.J., J.D. SIMMONS, J.D. GRANT,J.A. MAGNUSON and T.L. GOODFRIEND (1973) Clin. Chem. 19, 1339. UTIGER,R.D., M.L. PARKER and W.H. DAUGHADAY (1962) J. Clin. Invest. 41,254. VAITUKAITIS, J., J.B. ROBBINS, E. NIESCHLAG andG.T. Ross (1971) J. Clin. Endocrinol. Metab. 33, 988. VOLLER,A.. D.E. BIDWELL and A. B A R T L E T T (Bull. ~~~ Wld ~ ) Hlth Org. 53,55. WEEMEN, B.K. van and A.H.W.M. SCHUURS (1971) FEBS Lett. 15,232. and H.E. BOND (1977) Clin. Chem. 23, WEI, R., G.J. KNIGHT,D.H. ZIMMERMAN 813. WIDE,L. and J. PORATH (1966) Biochim. Biophys. Acta 130, 257. (1967) Lancet ii, 1105. WIDE,L., J. BENNICHand S.G.O. JOHANSSON WISDOM, G.B. (1976) Clin. Chem. 22, 1243. WOODHEAD, J.W., G.M. ADDISON and C.N. HALES(1974) Brit. Med. Bull. 30,44. YALOW, R.S. and S.A. BERSON (1960) J. Clin. Invest. 39, 1157. YALOW,R.S. and S.A. BERSON(1968) In: R.L. Hayer, F.A. Goswitz and B.E.P. Murphy (eds.) Radioisotopes in Medicine. In Vitro Studies (A.E.C. Symposium Series No. 11) p. 7. ~ In: W. Odell and W.H. Daughaday (eds.) YALOW,R.S. and S.A. B E R S O(1971) Principles of Competitive-binding Protein Analysis (Lippincott, Philadelphia, Pa.).
Subject index
Accuracy 479480 ACTH 336,342, 351, 366, 386, 396, 427, 430, 476477, 498 adjuvant 390-391 adsorption 366, 407-410, 428433, 474475 affinity constant 308-310 importance 321-327 measurement 322-327, 384, 394, 46446 5 alpha-fetoprotein 334, 386 alternative label 303-304, 374 et seq. ammonium sulphate 41041 1,472 androgens 434 anilino-naphthalene sulphonic acid 437438 antibodies absorption 471472 allergic 399400 availability 384, 393 characteristics 384-385, 393-396 chemistry 379-381 combining site 381-382,497498 heterogeneity 465 production of 385 et seq. selection of 393-396 specificity 384, 393 storage of 396 titre 394 to growth hormone 399,438 antigens 381-382 531
immunogenicity 386-388 assay mean 493 assay services 510 et seq. automation 504 et seq. Background counts 350 bacteriophage labels 376 beta-particles 344, 346, 524 between-assay variation 490 binder and sensitivity 450452, 458460 availability 378 characteristics 377 et seq. detection 398-400 specificity 378 binding assay 302-304 biological assays 303,495 blank values 402403,433. 473, 480. 493 ‘bridge’ antibodies 358 buffers 327-328 Calculation of results 313, 440 et seq. electronic 443445, 508-509, 520 linearisation 441443 manual 440-441 carcinoembryonic antigen 334 carry over 505-506 centrifuges 519 charcoal 341,408409,472 chemical precipitation 365 chloramine T 306, 353-354,358-360,423
532
RADIOIMMUNOASSAY A N D RELATED TECHNIQUES
coated tubes 418419, 508 coefficient of variation 490 collection of samples 475476 competitive protein binding 302-304, 397-398,496497 computers 443445,461462 confidence limits 445 conjugation labelling 355, 357 counters 345-350, 508, 513-514, 518 automatic 348 manual 348 outputs 3 4 9 , 4 4 3 4 5 counting errors 485487 counting time 348-350 crystal size 348 curie 345 Cusum plot 492493 Damage 360-364 decay catastrophe 363 dialysis 366 dilution curves 31 I dinitrophenol391 disequilibrium assays 452454, 470-471, 481, 507 dissociation constant 307 double antibody 413418 prozone 415416 solid phase 417 4 18 Electrophoresis 368, 370-371, 405-406, 519 enzymes 474-476,477, 500-501 enzymoimmunoassay 374-375 equilibrium 308 equipment 513-514, 518-520 extraction methods 427 et seq. recovery 43 1 reproducibility 431 specificity... 431433,434-436 Fibrinogen 336,471472, 501-503
Florisil 410, 430 fluoroimmunoassay 374 follicle-stimulating hormone 331, 333, 470,47 1 fragments 497-503 free hormones 438 free radical labels 376 Fuller’s earth 409. 430 Gamma-rays 344-347, 524 gel filtration chromatography 366, 406407, 507-508 glass beads 430 glucagon 336 growth hormone 331-332,373,399,438 Half-life 345, 347 haptens 382,385, 387, 388-390 heparin 473 hepatitis antigen 336, 375 heteroscedasticity 442 history 301-302,304306 hydroxyapatite 410 Immune response 382-384 immunisation 385 et seq. animal species 391-392 route 392-393 specificity 330,469470 timing 393 immunoassay 302-304 immunoglobulin 379 immunoradiometric methods 422426 incubation time 452454 insulin 305, 336,476 iodination 353 et seq. conjugation labelling 355 damage 360-364 electrolytic 355 vaporisation 355 I2%odine352, 353 et seq., 358, 373 monochloride 354
SUBJECT INDEX
ion-exchange chromatography 366 isotopes (radioactive) 304, 485487 chemistry 343-345 isotopic abundance 358 Kits 514-515, 522-523 Laboratory space 513 lactoperoxidase 355 late addition 454 Law of Mass Action 308 ligdnd 302 ligand-free fluids 476478 logit transform 320, 4 4 1 4 2 , 468, 485, 493 luteinising hormone 331, 333, 338, 470, 499-500 lymphocytes 382-384 lyophilisation 484, 489 Magnetic particles 419, 508 model system 324327 multiple detectors 349 myoglobin 501 Neo-antigen 471472 non-parallelism 432, 464465, 476 Oestrogens 361, 389-390, 396, 400, 435, 436 oxytocin 339,427,478,498499 Parathormone 336 peptides 335 pipettes 519 placental lactogen 315-318, 332, 340, 366-367, 373. 385,386,458 polyacrylamide gel 41 9 polyethylene glycol 317,411412,413 precision 479 et seq. definition 479480 errors in assay 480-488
533
principles 306 progesterone 434 prolactin 331-332, 367, 373 prostaglandins 352, 390, 500 protein hormones 333-334 purification antibodies 454456 iodination mixtures 365 et seq. proteins 334 Quality control 488494 quenching 347 Quso 409 Radioactivity 344-345, 523-526 reagent supply 514-515, 520-522 receptor assay 302-304, 396-397, 400, 464,496 reproducibility 479480 restricted population mean 494 Safety precautions 523-525 sample identification 505 scales, arithmetic and logarithmic 321 Scatchard plot 322, 395 scintillators 345-348 second antibody 413418 prozone 41 5 4 1 6 solid phase 417418 sensitivity concentration of sample 427429 definition 446448 optimisation 325-327,448462 separation methods 401 et seq. adsorption 407410 and precision 480482 and specificity 472 automated 507-508 efficiency 4014 0 3 electrophoretic 405406 fractional precipitation 410413 gel filtration 406407
534
RADIOIMMUNOASSAY A N D RELATED TECHNIQUES
immunoradiometric 422426 practicality 403404 second antibody 413418 solid phase 418421 sex hormone binding globulin 397, 399, 434,497 silicates 409410 solid-phase systems 418421,472, 508 specific activity 347, 353, 370, 372-374, 449,458 specificity 463 et seq. assessment 466499 definition 466467, 473 non-specific non-specificity 472478 specific non-specificity 464-472 staff 511-513 standard curves 31 1-321,484-485 standard curve methods of plotting 3 1 4 321,440443 standards 310-311, 321, 337-341, 481484 steroid hormones 334-335, 356, 367, 386387,434436 storage 341-342,480 supraregional assay service 515-517 Talc 409 targeting of assays 457461 technical skill 480,487488, 504,512-513
temperature effects 456-457 terminology 302-304 thin-layer chromatography 367,478 thyroid-stimulating hormone 333 thyroxine 305,437,468.476 thyroxine binding globulin 305, 397-398, 399,437438 tracer 303, 343 and sensitivity 449450,458 assessment 369 et seq. characteristics 351 damage 35 I external 351-352 internal 351-352 purification 351 et seq. triiodothyronine 468 tritium 344 two-site assays 424-425 tyrosine 353 Vasopressin 339, 352, 362,427, 432 vitamin BIZ305 Wick chromatography 371-372 Wilzbach technique 352 within-assay variation 490 Yield 370